Virtual camera for 3D maps
Through integrated map drawing applications and virtual camera technology, 3D map tiles are generated and presented, which solves the problem of low 3D map processing efficiency in the existing technology, and realizes efficient 3D map rendering and display, improving user experience and application performance .
Patent Information
- Application Number
- CN202110507529.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2012-09-30
- Filing Date
- 2013-05-31
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2033-05-31
AI Technical Summary
In the prior art, map-based applications are inefficient when processing 3D maps, slow and unformatted applications, resulting in poor user experience and difficult to achieve smooth transitions of immersive and non-immersive 3D maps.
Using integrated map drawing applications, the 3D display is rendered using the location of the virtual camera, and a smooth transition between immersive and non-immersive displays are achieved on the client device, using the image processing pipeline. Preload and decompress operations, optimize rendering pipelines for efficiency.
It realizes efficient 3D map rendering and display, improves user experience, and can smoothly transition on the device. Immersive and non-immersive displays, improving application performance and user interactivity.
Smart Images

Figure CN113155147B_ABST
Abstract
Description
[0001] This application is a divisional application of the invention patent application with application date of May 31, 2013, application number 201380028723.6, and invention name “Virtual Camera for 3D Maps”. Background Art
[0002] Many map-based applications available today are designed for use on a variety of devices (e.g., desktop computers, laptops, tablets, smartphones, handheld Global Positioning System (GPS) receivers, etc.) and for a variety of purposes (e.g., navigation, browsing, sports, etc.). Most of these applications generate a display of a map based on map data that describes the relative positions of streets, roads, points of interest, etc. in the map.
[0003] The maps used in such applications are typically two-dimensional (2D) or three-dimensional (3D). However, a large number of applications use 2D maps, partly due to the processing-intensive requirements of viewing 3D maps. For the same reasons, applications that use 3D maps are often slow, inefficient, unformatted, and / or simplistic, rendering the applications useless. Summary of the Invention
[0004] Some embodiments of the present invention provide novel methods for generating and presenting immersive and non-immersive 3D map displays for browsing and navigation. In some embodiments, the immersive and non-immersive 3D map displays can be displayed alternately and / or sequentially on a device (e.g., a mobile device) having a touch-sensitive screen and a multi-touch interface that allows a user to interact with the display through touch and gesture input on the screen. In other embodiments, the display is provided on a device that does not have a touch-sensitive screen.
[0005] In some embodiments described below, these 3D displays are provided by an integrated mapping application that provides several useful modalities, such as location browsing, map search, route identification, and route navigation operations. However, in other embodiments, the mapping application does not utilize all of these modalities. For example, in some embodiments, the mapping application does not provide route navigation.
[0006] In order to display both immersive and non-immersive 3D map presentations, some embodiments must generate various tiles for rendering by the client device to generate roads, buildings, and surrounding scenery. In some embodiments, examples of such tiles include road and building tiles for non-immersive 3D presentations, and navigation and building tiles for immersive 3D presentations.
[0007] Before generating these tiles, a set of servers must generate descriptions of the roads, buildings, and other geometric structures that are placed in each tile. This task involves multiple subtasks, such as (1) receiving map data from various vendors, (2) processing such data to produce one-dimensional (1D) roads, (3) smoothing the 1D road map, (4) defining data to specify junction points, (5) generating 2D road geometry and land cover, (6) smoothing the 2D road geometry, (7) generating data about buildings (e.g., estimated height data), (8) using such data to define building geometry, (9) constructing road geometry details (such as islands, lane markings, and distances between road geometry and land cover), and (10) identifying geometry edge node features and propagating such features.
[0008] Once the tiles are generated on the mapping service side, the tiles are used by the mapping application on the client device to present the 3D map to the user of the client device. The mapping application of some embodiments uses various novel techniques to present a 3D presentation of the map while the map is being browsed or while the map is providing a navigation presentation. For example, the mapping application renders the 3D presentation based on the vantage point of the virtual camera, and uses various methods of automatically moving the virtual camera (i.e., moving the perspective rendering position) in certain situations to achieve the desired 3D presentation. One such instance occurs when the mapping application of some embodiments moves the position of the virtual camera from a perspective rendering position behind the current position of the moving device to a top-down view of the current position as the device is about to turn a corner along a route. Another example is the zoom in / out animation provided that shows objects in the scene getting larger and shrinking as the zoom in and zoom out operations occur.
[0009] In addition, in some embodiments, the mapping application provides two different types of 3D presentations: an immersive 3D presentation and a non-immersive 3D presentation. The immersive presentation in some embodiments displays not only more geometry but also more detail than the geometry displayed in the non-immersive presentation. The mapping application also provides a smooth transition between the non-immersive presentation and the immersive presentation.
[0010] To achieve such smooth transitions and generate other novel effects, the mapping application of some embodiments uses a novel image processing pipeline. This pipeline performs various pre-loading operations to download, retrieve, and / or decompress map tiles that may be needed for navigation presentations, prepares its rendering pipeline for its rendering operations, and prepares a replication pipeline to smoothly transition between immersive and non-immersive 3D presentations.
[0011] The above summary of the invention is intended to serve as a brief introduction to some embodiments of the present invention. It is not meant to introduce or summarize all the subject matter disclosed in this document. The subsequent detailed description and the drawings referenced in the detailed description will further describe the embodiments described in the summary of the invention as well as other embodiments. Therefore, in order to understand all the embodiments described in this document, a comprehensive review of the summary of the invention, the detailed description and the drawings is required. In addition, the subject matter protected by the claims is not limited by the exemplary details in the summary of the invention, the detailed description and the drawings, but is limited by the appended claims, because the subject matter protected by the claims can be embodied in other specific forms without departing from the essence of the subject matter. BRIEF DESCRIPTION OF THE DRAWINGS
[0012] The novel features of the invention are set forth in the appended claims.For illustrative purposes, however, several embodiments of the invention are set forth in the following drawings.
[0013] Figure 1 An example of a device that executes an integrated mapping application of some embodiments of the present invention is shown.
[0014] Figure 2 It shows how the navigation application of some embodiments provides 3D controls as a quick mechanism to enter 3D navigation mode.
[0015] Figure 3 A simplified example is presented to illustrate the concept of a virtual camera. Figure 4 Adjustment of the distance of the virtual camera by pinching and expanding gestures is shown.
[0016] Figure 5 An embodiment of a camera whose angle can be adjusted by hand gestures is shown.
[0017] Figure 6 The mapping service processing of some embodiments is conceptually illustrated.
[0018] Figure 7 The process of some embodiments for generating a road map is conceptually illustrated.
[0019] Figure 8 Various data structures of some embodiments are shown for roads within a map area.
[0020] Figure 9 An example of modification for two road segments meeting at a junction is shown.
[0021] Figure 10 Shown is an example of three road segments that are candidates for aggregation into a road at a joining point.
[0022] Figure 11The situation where the first road segment intersects (in a plane) with the second road segment and the third road segment is shown.
[0023] Figure 12 Two sections are shown that are a distance X apart, extend parallely in opposite directions, and have the same name.
[0024] Figure 13 and Figure 14 A similar winding road is shown with speed limits of 25 mph and 60 mph.
[0025] Figure 15 Conceptually illustrated are processes of some embodiments for generating road geometry to be used to generate road polygons on a client mapping application.
[0026] Figure 16 The operation of performing a smooth lane extension join is shown, wherein one of the two sides of the road remains straight (ie, is "aligned").
[0027] Figure 17 Shown from Figure 16 Result of smooth lane extension as rendered by the client-side mapping application of some embodiments.
[0028] Figure 18 The operation of performing a smooth lane extension join is shown, where the road is extended on both sides.
[0029] Figure 19 Shown from Figure 18 Result of smooth lane extension as rendered by the client-side mapping application of some embodiments.
[0030] Figure 20 A tapering operation is shown that is performed to smooth corners between road segment geometries when the angle between the road segments is greater than a first threshold angle.
[0031] Figure 21 Shown from Figure 20 The result of smooth joins as rendered by the client-side mapping application of some embodiments.
[0032] Figure 22 A projection operation is shown performed to smooth corners between road segment geometries when the angle between the road segments is less than a first threshold angle and greater than a second threshold angle.
[0033] Figure 23 Shown from Figure 22 The result of smooth joins as rendered by the client-side mapping application of some embodiments.
[0034] Figure 24A clipping operation is shown that is performed to eliminate redundant roads of the road segment geometry when the angle between the road segments is less than a first threshold angle and greater than a second threshold angle.
[0035] Figure 25 An intersection operation is shown that is performed to smooth corners between road segment geometries.
[0036] Figure 26 Shown from Figure 24 and Figure 25 The result of smooth joins as rendered by the client-side mapping application of some embodiments.
[0037] Figure 27 Shown is a tapering operation performed in some embodiments to smooth corners between road segment geometries in situations that conform to the characteristics of a freeway on-ramp.
[0038] Figure 28 Shown from Figure 27 Result of a smooth highway merge junction as rendered by a client-side mapping application of some embodiments.
[0039] Figure 29 The generation of additional geometry at the join point is shown to create a more realistic rounded corner at the join point.
[0040] Figure 30 Shown from Figure 29 The result of a smooth intersection as rendered by a client-side mapping application of some embodiments.
[0041] Figure 31 It shows that two roads are pushed apart when they overlap.
[0042] Figure 32 It shows that when the road segments overlap, the width of both road segments is reduced.
[0043] Figure 33 and Figure 34 A dual carriageway is shown with two different types of central dividers.
[0044] Figure 35 An example of the geometry of a junction of two main arterial roads is shown.
[0045] Figure 36 Shown from Figure 35 The result of the join points as rendered by the client-side mapping application of some embodiments.
[0046] Figure 37 An example of using vertex annotations to specify the bridge frame of an interchange is shown, in which one road runs over another road without the roads forming a junction.
[0047] Figure 38 Shown Figure 37 The result of the road segment as rendered by the client mapping application of some embodiments.
[0048] Figure 39 An example of using vertex annotations to designate edges of a road segment geometry as inner edges is shown, indicating that boxes should not be drawn for guiding edges from the annotated vertex.
[0049] Figure 40 Shown Figure 39 The result of the road segment as rendered by the client mapping application of some embodiments.
[0050] Figure 41 An example of a tunnel for a road segment is shown using vertex annotations to indicate that a tunnel should be drawn on the road segment.
[0051] Figure 42 Shown Figure 41 The result of the road segment as rendered by the client mapping application of some embodiments.
[0052] Figure 43 An example of using vertex annotations to specify a sidewalk on one side of a road segment is shown.
[0053] Figure 44 Shown Figure 43 The result of the road segment as rendered by the client mapping application of some embodiments.
[0054] Figure 45 An example is shown using several annotations for a single vertex of some embodiments.
[0055] Figure 46 Shown Figure 45 The result of the road segment as rendered by the client mapping application of some embodiments.
[0056] Figure 47 An example of using vertex annotation data to change the properties of a road box of the road geometry is shown.
[0057] Figure 48 Shown Figure 47 The result of the road segment as rendered by the client mapping application of some embodiments.
[0058] Figure 49 Conceptually illustrates the operations performed by a mapping service of some embodiments to generate a route for a requesting device and provide the route along with navigation instructions to the requesting device.
[0059] Figure 50Conceptually illustrates a process performed by the mapping service of some embodiments to generate and transmit route and intersection data to a user.
[0060] Figure 51 The process of some embodiments for determining path segments between a set of join points that should be treated together as a single intersection is conceptually illustrated.
[0061] Figure 52 Examples of junctions of some embodiments are shown and illustrate that it is not required that path segments meet at right angles or that the path continue in a straight line through the junction.
[0062] Figure 53 An intersection of some embodiments is shown including two dual carriageway paths and a one-way street.
[0063] Figure 54 The process of some embodiments for joining several junctions together into a single intersection and identifying branches of the intersection is conceptually illustrated.
[0064] Figure 55 Some embodiments illustrate a co-existing intersection between a dual carriageway with two paths and a dual carriageway with two paths.
[0065] Figure 56 An intersection of some embodiments is shown where the left turn lane is defined as a separate path segment.
[0066] Figure 57 A switch in an intersection of some embodiments is shown.
[0067] Figure 58 An intersection with both a divergence and a left-turn lane is shown for some embodiments.
[0068] Figure 59 Additional bidirectional paths in intersections are shown in some embodiments.
[0069] Figure 60 The reduction of an eight-path intersection to four branches is shown in some embodiments.
[0070] Figure 61 Shown is the reduction of a different eight-path intersection into five branches in some embodiments.
[0071] Figure 62 An example of a data structure for a point-type intersection according to some embodiments is conceptually illustrated.
[0072] Figure 63 The data structure of some embodiments for a roundabout is conceptually illustrated.
[0073] Figure 64 Conceptually illustrated is the reduction of a roundabout to intersection data in some embodiments.
[0074] Figure 65 The process of some embodiments for modifying intersection data to provide navigation information for a route is conceptually illustrated.
[0075] Figure 66 A conceptual rendering of a route through an intersection, a data structure for the intersection, and modifications to the data structure to create a new data structure for turn-by-turn navigation instructions are shown.
[0076] Figure 67 A process performed by some embodiments of the present invention for merging land cover polygons into road polygons is conceptually illustrated.
[0077] Figure 68 An example of merging land cover into road polygons is shown.
[0078] Figure 69 Conceptually illustrates a process that, in some embodiments, merges intersections between adjacent polygons received from two different sources.
[0079] Figure 70 Two stages of some embodiments of the present invention are shown that resolve identified gaps by merging boundary coordinates between adjacent polygons.
[0080] Figure 71 Two stages of some embodiments of the present invention are shown that merge the boundary coordinates around identified overlapping polygons.
[0081] Figure 72 A process performed by some embodiments of the present invention for resolving boundary conflicts between regions is conceptually illustrated.
[0082] Figure 73 An example of resolving a boundary conflict between two adjacent regions is shown.
[0083] Figure 74 Conceptually illustrates an example of a virtual camera automatically moving around an area of a 3D map based on route navigation according to some embodiments of the present invention.
[0084] Figure 75 The process of some embodiments for animating a map view during route navigation is conceptually illustrated.
[0085] Figure 76Conceptually illustrates an example of a virtual camera automatically maneuvering turns in a 3D map during route navigation according to some embodiments of the present invention.
[0086] Figure 77 A process of some embodiments for animating a map view during turns in route navigation is conceptually illustrated.
[0087] Figure 78 Conceptually illustrated is an example of a virtual camera panning around a 3D map during route navigation according to some embodiments of the present invention.
[0088] Figure 79 The process of some embodiments for panning to the side of a map view based on gesture input is conceptually illustrated.
[0089] Figure 80 Conceptually illustrates the perspective adjustment features provided by the mapping application of some embodiments.
[0090] Figure 81 Conceptually illustrates a process of some embodiments for adjusting the position of a virtual camera used to render a map view in response to a viewing angle adjustment.
[0091] Figure 82 A virtual camera is conceptually illustrated that can be moved along different arcs and at different zoom levels according to some embodiments of the present invention.
[0092] Figure 83 The process of some embodiments for determining an arc along which a virtual camera may move is conceptually illustrated.
[0093] Figure 84 Conceptually illustrates a feature provided by a mapping application of some embodiments for maintaining the position of a virtual camera within a defined range along an arc.
[0094] Figure 85 A zoom adjustment feature provided by a mapping application of some embodiments is shown.
[0095] Figure 86 Conceptually illustrates the process of some embodiments of adjusting the position of a virtual camera used to render a map view in response to a zoom level adjustment.
[0096] Figure 87 A rotation operation of a 3D map view according to some embodiments of the present invention is conceptually illustrated.
[0097] Figure 88 Conceptually illustrates a rotation operation of a 3D map view after activating a location indicator feature, according to some embodiments of the present invention.
[0098] Figure 89 Conceptually illustrates a rotation operation of a 3D map view from a 2D perspective of the 3D map according to some embodiments of the present invention.
[0099] Figure 90 The process of some embodiments for rotating a map view based on gesture input is conceptually illustrated.
[0100] Figure 91 Conceptually illustrated is an example of a mapping application of some embodiments that renders 3D map views for different zoom levels for different viewing modes.
[0101] Figure 92 Movement of a virtual camera around an area of a 3D map defined as viewed by the immersive viewing mode of some embodiments is shown.
[0102] Figure 93 Movement of a virtual camera around an area of a 3D map defined as viewed by a non-immersive viewing mode of some embodiments is shown.
[0103] Figure 94 Conceptually illustrated is an example of a mapping application of some embodiments that renders 3D map views for different zoom levels for different viewing modes.
[0104] Figure 95 Conceptually illustrates a processing pipeline that is performed by a mapping application of some embodiments for rendering a 3D map view based on different 3D map tiles.
[0105] Figure 96 Data structures used by the processing pipeline of some embodiments are conceptually illustrated.
[0106] Figure 97 A state diagram of a mapping application of some embodiments is conceptually illustrated.
[0107] Figure 98 Conceptually illustrates a processing pipeline that is performed by a mapping application of some embodiments to render a map for display at a client device.
[0108] Figure 99 The application of different textures to different types of roads is shown for some embodiments.
[0109] Figure 100 A map representation including one main road and two connecting roads is shown.
[0110] Figure 101A non-immersive map presentation is shown showing roads rendered in map tiles in a map browsing view.
[0111] Figure 102 Conceptually illustrated is a controller instantiating two virtual cameras for two different views of a map area.
[0112] Figure 103 The process of some embodiments for preloading navigation tiles for entry navigation is conceptually illustrated.
[0113] Figure 104 A mapping application is shown as an application transitioning from a non-immersive map view for map browsing to an immersive map view for navigation.
[0114] Figure 105A and Figure 105B State diagram 10500 is conceptually shown, describing the different states of the integrated mapping, search, and navigation application of some embodiments and the transitions between these states.
[0115] Figure 106 An example of an architecture for a mobile computing device.
[0116] Figure 107 An example of an electronic system with which some embodiments of the present invention may be implemented is conceptually illustrated.
[0117] Figure 108 A mapping service operating environment according to some embodiments is shown. DETAILED DESCRIPTION
[0118] Some embodiments of the present invention provide novel methods for generating and presenting immersive and non-immersive 3D map displays for browsing and navigation. In some embodiments, the immersive and non-immersive 3D map displays can be displayed alternately and / or sequentially on a device (e.g., a mobile device) having a touch-sensitive screen and a multi-touch interface that allows a user to interact with the display through touch and gesture input on the screen. In other embodiments, the display is provided on a device that does not have a touch-sensitive screen.
[0119] In some embodiments described below, these 3D displays are provided by an integrated mapping application that provides several useful modalities, such as location browsing, map search, route identification, and route navigation operations. However, in other embodiments, the mapping application does not utilize all of these modalities. For example, in some embodiments, the mapping application does not provide route navigation.
[0120] In order to display both immersive and non-immersive 3D map presentations, some embodiments must generate various tiles for rendering by the client device to generate roads, buildings, and surrounding scenery. In some embodiments, examples of such tiles include road and building tiles for non-immersive 3D presentations, and navigation and building tiles for immersive 3D presentations.
[0121] Before generating these tiles, a set of servers must generate descriptions of the roads, buildings, and other geometric structures placed in each tile. This task involves multiple subtasks, such as (1) receiving map data from various vendors, (2) processing such data to produce one-dimensional (1D) roads, (3) smoothing the 1D road map, (4) defining data to specify junction points, (5) generating 2D road geometry and land cover, (6) smoothing the 2D road geometry, (7) generating data about buildings (e.g., estimated height data), (8) using such data to define building geometry, (9) constructing road geometry details (such as islands, lane markings, and distances between road geometry and land cover), and (10) identifying geometry edge node features and propagating such features.
[0122] Once the tiles are generated on the mapping service side, the tiles are used by the mapping application on the client device to present the 3D map to the user of the client device. The mapping application of some embodiments uses various novel techniques to present a 3D presentation of the map while the map is being browsed or while the map is providing a navigation presentation. For example, the mapping application renders the 3D presentation based on the vantage point of the virtual camera, and uses various methods of automatically moving the virtual camera (i.e., moving the perspective rendering position) in certain situations to achieve the desired 3D presentation. One such instance occurs when the mapping application of some embodiments moves the position of the virtual camera from a perspective rendering position behind the current position of the moving device to a top-down view of the current position as the device is about to turn a corner along a route. Another example is the zoom in / out animation provided that shows objects in the scene getting larger and shrinking as the zoom in and zoom out operations occur.
[0123] In addition, in some embodiments, the mapping application provides two different types of 3D presentations: an immersive 3D presentation and a non-immersive 3D presentation. The immersive presentation in some embodiments displays not only more geometry but also more detail than the geometry displayed in the non-immersive presentation. The mapping application also provides a smooth transition between the non-immersive presentation and the immersive presentation.
[0124] To achieve such smooth transitions and generate other novel effects, the mapping application of some embodiments uses a novel image processing pipeline. This pipeline performs various pre-loading operations to download, retrieve, and / or decompress map tiles that may be needed for navigation presentations, prepares its rendering pipeline for its rendering operations, and prepares a replication pipeline to smoothly transition between immersive and non-immersive 3D presentations.
[0125] The following Section I describes mapping applications of some embodiments of the present invention. Section II then describes server-side operations for generating tiles required to produce immersive 3D presentations and non-immersive 3D presentations. Section III then describes client-side operations for generating immersive 3D presentations and non-immersive 3D presentations. Section IV then describes electronic devices using the mapping applications of some embodiments. Section V finally describes location services used by some embodiments of the present invention.
[0126] In the following detailed description of the invention, many details, examples and embodiments of the present invention are proposed and described. However, it will be clear and apparent to those skilled in the art that the present invention is not limited to the embodiments shown, and the present invention can be practiced without the specific details and examples discussed.
[0127] I. Navigating the User Interface
[0128] A. Start
[0129] The navigation application of some embodiments is part of an integrated mapping application that includes several useful modalities, including location browsing, map searching, route identification, and route navigation operations. The integrated application in some embodiments (hereinafter referred to as the mapping application, navigation application, or integrated application) is defined to be executed by a device having a touch-sensitive screen that displays the output of the application. In some embodiments, the device has a multi-touch interface for allowing a user to interact with the application by providing touch and gesture input through the screen. Examples of such devices are smartphones (e.g., the iPhone sold by Apple Inc. operate Mobile phone operating system, operating Windows operating system, etc.).
[0130] Figure 1An example of a device 100 that performs an integrated mapping application of some embodiments of the present invention is shown. The figure also shows an example of initiating route navigation in the application. The application has a novel user interface (UI) design that seamlessly and harmoniously integrates controls for each of the different modes of the application by using a minimal set of on-screen controls that float on top of the content so that as much content as possible is displayed. In addition, the cluster adapts to the task at hand, animating the contents of the cluster as the user moves between different modes (e.g., between browsing, searching, routing, and navigating). The adaptive nature of this common element enables the mapping application to be optimized for different tasks while maintaining a consistent appearance and interaction model when moving between those tasks.
[0131] Figure 1 Six stages 105, 110, 115, 117, 119, 121 of interaction with a mapping application are shown. The first stage 105 shows a UI 120 of the device, which includes several icons for several applications in a taskbar area 125 and on one page of the UI. One of the icons on the page is an icon for a mapping application 130. The first stage shows user selection of the mapping application by touch contact with the device screen at the location of the application on the screen.
[0132] The second stage 110 shows the device after the mapping application is opened. As shown in this stage, the UI of the mapping application has a start page that, in some embodiments, (1) displays a map of the device's current location, and (2) several UI controls arranged in a top bar 140 and as floating controls. Figure 1 As shown in FIG, the floating controls include an indicator 145, a 3D control 150, and a scroll control 155, while the top bar 140 includes a direction control 160, a search field 165, and a bookmark control 170.
[0133] In some embodiments, the user can initiate a search by tapping in the search field 165. This directs the app to present an animation that (1) presents an on-screen keyboard and (2) opens a search form filled with the ever-present completion items. This form has some important subtleties. When the search field is tapped and before editing a word, or when the search field is empty, the form contains a list of "recent items," which in some embodiments are the most recent searches and directions the user has requested. This makes it very easy to quickly bring up recently accessed results.
[0134] After any edits in the search field, the form is filled with search completions from both local sources (e.g., bookmarks, contacts, recent searches, recent directions, etc.) and remote servers. Incorporating the user's contact card into the search interface adds additional flexibility to the design. When the recent items are shown, in some embodiments, directions from the current location to the user's home are always provided, while in other embodiments the directions are provided in a context deemed "appropriate." Additionally, when the search term matches at least a portion of an address tag (e.g., "Wo" or "ork" for "Work"), in some embodiments the application presents the user's marked address as a completion in the search form. Together, these behaviors make the search UI a very powerful way to put results from various sources on a map. In addition to allowing the user to initiate a search, the presence of a text field in the main map view in some embodiments also allows the user to view the query corresponding to the search results on the map and remove those search results by clearing the query.
[0135] A bookmark control 170 (e.g., a button) allows locations and routes to be bookmarked by the application. A location indicator 145 allows the device's current location to be specifically marked on the map. Once the indicator is selected once, the application maintains the device's current location at the center of the map. In some embodiments, the location indicator can also identify the direction the device is currently pointing.
[0136] The 3D control 150 is a control for viewing a map or checking a route in three dimensions (3D). The mapping application provides the 3D control as a quick mechanism for entering and exiting 3D. The control also serves as an indicator of (1) that the current view is a 3D view, (2) that a 3D perspective is available for a given map view (e.g., a zoomed-out map view may not have a 3D view available), (3) that a 3D perspective is not available (e.g., 3D data is not available for a map area), and (4) that flyover animations are available at a given zoom level. The 3D control may provide a different appearance corresponding to each indication. For example, the 3D control may be gray when a 3D view is not available, black when a 3D view is available but the map is in a 2D view, and blue when the map is in a 3D view. In some embodiments, the 3D control changes to an image of a building when flyover animations are available for the user's given zoom level and location on the map.
[0137] The scroll control 155 is a control that allows an application to minimize the number of controls on the screen by placing certain less frequently used actions in a secondary UI screen that can be accessed via a "scroll" control displayed on the map. In some embodiments, the scroll is permanently displayed on at least some of the map views provided by the application. For example, in some embodiments, the application permanently displays the scroll on the start page (shown in the second stage 110) provided by the application to allow the user to browse or search for a location or identify a route.
[0138] Directions control 160 opens a directions entry page 180 through which a user may request a route to be identified between a starting location and an ending location. Figure 1 The third stage 115 shows that selection of the directions control 160 opens a directions entry page 180, which is shown in the fourth stage 117. The directions control is one of three mechanisms by which a mapping application can be directed to identify and display a route between two locations; the other two mechanisms are (1) a control in an information banner displayed for an item selected in the map, and (2) the most recent routes identified by the device displayed in a search field 165. Thus, the information banner control and the search field 165 are two UI tools used by the application to make transitions between different modalities seamless.
[0139] Stage 4 117 shows a directions entry page 180 including start and end fields for providing the starting and ending locations of a route, and a table listing the most recent routes that the application has provided to the user. Other controls on the page are a control for starting a route, a control for reversing the order of the starting and ending locations, a control for canceling a directions request, and a control for selecting a walking, car, or public transportation route. These controls and other aspects of the mapping application are described in the following provisional patent applications and patent applications: U.S. Provisional Patent Application 61 / 656,080, entitled “Integrated Location Browsing, Map Searching, Route Identifying, and Route Navigating Application,” filed on June 6, 2012; U.S. Provisional Patent Application 61 / 699,841, entitled “Problem Reporting,” filed on September 11, 2012; and U.S. Patent Application 13 / 632,102, entitled “Problem Reporting in Maps,” filed on September 30, 2012. Provisional applications 61 / 656,080 and 61 / 699,841, and patent application 13 / 632,102 are incorporated herein by reference.
[0140] The fourth stage shows the user selecting one of the most recent directions automatically populated in table 182. The fifth stage 119 then shows three routes on the 2D map view between the specified start and end locations specified via page 180. The fifth stage also shows the selection of a second route and some information about this route in a bar at the top of the layout. The bar is shown as including a start button and an end button. The start button is shown as being selected in the fifth stage.
[0141] As shown by the sixth stage, selection of the Start button instructs the application to enter turn-by-turn navigation mode. In this example, the application has entered 2D turn-by-turn navigation mode. In other embodiments, the application will default to 3D turn-by-turn navigation mode. In this mode, the application displays a realistic sign 184 that identifies the distance from the device's current location to the next maneuver in the navigated route, as well as some other relevant information. The application also displays a top bar that includes some information about the navigation, as well as an End button and an Overview button for ending the navigation and obtaining the remaining portion of the navigated route, or in other embodiments, obtaining an overview of the entire portion of the navigated route, respectively.
[0142] The mapping application of some embodiments uses coordinates (e.g., longitude, altitude, and latitude) in GPS signals received by the device at the device's location to identify the device's location. Alternatively, or in combination, the mapping application uses other methods (e.g., cell tower triangulation) to calculate the current location. When a user carrying the device deviates from a route, the mapping application of some embodiments tracks the device's location and recalculates a new route based on the deviated location to redirect the user from the deviated location to the target location. In other words, the mapping application of some embodiments operating in navigation mode requires that the device always be on the route.
[0143] The application further displays the floating 3D control and floating list control described above. It should be noted that upon entering route check and route navigation modes, the list control is adaptively added to the floating control cluster, while upon entering route navigation mode, the location indicator is removed from the floating control. Additionally, upon transitioning from route check mode to route navigation mode, in some embodiments, the application performs an animation that involves fully expanding the scroll bar before the application transitions to the navigation display.
[0144] In some embodiments, the animated transition includes removing the top bar, controls associated with the top bar, and floating controls from the navigation display, and moving the indicator 184 to the top edge of the display a short time after the navigation display begins. In some embodiments, the application requires the user to tap the navigated map to restore the top bar, controls associated with the top bar, and floating controls, and requires another tap to remove these controls from the map again. Other embodiments provide other mechanisms for viewing and removing these controls. Other aspects of the navigation user interface and navigation mode of some embodiments are described in more detail in the following provisional patent applications and patent applications: U.S. Provisional Patent Application No. 61 / 655,997, titled “Route Navigating Method and Apparatus,” filed on June 5, 2012; U.S. Provisional Patent Application No. 61 / 657,864, titled “Route Navigating Method and Apparatus,” filed on June 10, 2012; U.S. Provisional Patent Application No. 61 / 699,851, titled “Voice Instructions During Navigation,” filed on September 11, 2012; and U.S. Patent Application No. 13 / 632,121, titled “Context-Aware Voice Guidance,” filed on September 30, 2012. Provisional Applications Nos. 61 / 655,997, 61 / 657,864, and 61 / 699,851, and Patent Application No. 13 / 632,121 are incorporated herein by reference.
[0145] B.2D Navigation and 3D Navigation
[0146] The navigation application of some embodiments may display a map for navigation in 2D mode or 3D mode.As mentioned above, one of the floating controls is a 3D control 150 that allows the user to view the navigation presentation in three dimensions (3D). Figure 2 It shows how the navigation application of some embodiments provides a 3D control 150 as a quick mechanism to enter 3D navigation mode. The figure shows this operation in three stages 205 - 215. The first stage 205 shows the user selecting the 3D control 150 while viewing a two-dimensional navigation presentation.
[0147] The second stage 210 shows the navigation display in the process of transitioning to a 3D display. As shown in the figure, the 3D control is highlighted at this stage to indicate that the navigation display has entered 3D mode. In some embodiments, the navigation application generates a 3D view of the navigated map by rendering the map view from a specific location in the three-dimensional scene, which can be conceptually regarded as the location of a virtual camera capturing the map view. Figure 3 This rendering is further described.
[0148] The third stage 215 then shows the navigation display at the end of its transition to its 3D appearance. As shown by the difference between the heights of the buildings in the second and third stages, the transition from 2D navigation to 3D navigation includes, in some embodiments, an animation showing three-dimensional objects in the navigated map becoming larger and larger.
[0149] 1. Virtual Camera
[0150] The navigation application of some embodiments can display a navigation map from multiple perspectives. The application can show the map in three dimensions (3D) or in two dimensions (2D). A 3D map is a generated simulation of a virtual scene as seen by a virtual camera. Figure 3 A simplified example is presented to illustrate the concept of a virtual camera 312. When rendering a 3D navigation map, a virtual camera is a conceptualization of a location in the 3D map scene from which the device renders a 3D view of the scene. Figure 3 A location in a 3D navigation map scene 310 including four objects, namely two buildings and two intersecting roads, is shown. To illustrate the virtual camera concept, the figure shows three scenarios, each of which corresponds to a different virtual camera position (i.e., a different rendering position) and a different resulting view displayed on the device.
[0151] The first stage 301 shows a virtual camera 312 pointing downward at a 3D scene 310 at a first position and at an angle (e.g., a 30° angle). By rendering the 3D scene according to the position and angle shown in stage 301, the application generates a 3D map view 318. Based on this position, the camera is pointing at the position of the moving position in front of the device. The virtual camera 312 remains behind the current position of the device. "Behind the current position" in this case means going in the opposite direction of the defined path of the navigation application, which is in the opposite direction of the current direction in which the device is moving.
[0152] Navigation map view 318 looks like it was taken by a camera from above and behind the device's location indicator 316. The position and angle of the virtual camera place location indicator 316 near the bottom of navigation map view 318. This also results in most of the screen being filled with streets and buildings in front of the device's current location. In contrast, in some embodiments, location indicator 316 is in the center of the screen, with half of the screen representing things in front of the device and the other half representing things behind the device. To simplify the figure, no road signs are shown for views 318, 328, and 338.
[0153] The second stage 302 shows a virtual camera 312 pointing downward at a second, larger angle (e.g., 45°) at a different location toward the scene 310. The application renders the scene 310 according to this angle, resulting in a 3D navigation map view 328. The buildings and roads are smaller than they were in the first navigation map view 318. The virtual camera 312 is once again positioned above and behind the location indicator 326 in the scene 310. This again causes the location indicator to appear in the lower portion of the 3D map view 328. The camera's position and orientation also once again causes most of the screen to display what is in front of the car, which is what the person navigating needs to understand.
[0154] The third stage 303 shows a virtual camera 312 in a top-down view looking down at a position on the 2D map 345 corresponding to a position in the 3D map scene 310 used to render the 3D views 318 and 328. The scene rendered according to this perspective is the 2D map view 338. Unlike the 3D rendering operations of the first and second stages, which in some embodiments are perspective 3D rendering operations, the rendering operations in the third stage are relatively simple because they only require cropping a portion of the 2D map identified by the zoom level specified by the application or user. Therefore, the description of the virtual camera characteristics in this case somewhat unnecessarily complicates the description of the application's operations because cropping a portion of the 2D map is not a perspective rendering operation.
[0155] At stage 303, when the camera switches from a 3D perspective view to a 2D top-down view, the mapping application in some embodiments switches from rendering the 3D scene according to a particular viewing direction to a cropped 2D scene. This is because in these embodiments, the application is designed to use simplified rendering operations that are easier and do not generate unnecessary viewing artifacts. However, in other embodiments, the mapping application uses viewing rendering operations to render the 3D scene according to a top-down virtual camera position. In these embodiments, the resulting 2D map view is slightly different from the map view 338 shown in stage 303 because any objects far from the center of the view are distorted, and the distortion increases the further away the objects are from the center of the view.
[0156] The virtual camera 312 moves along different trajectories in different embodiments. Figure 3 . Two such trajectories, 350 and 355, are shown in FIG. In both trajectories, the camera moves in an arc and rotates more downward as the camera moves up the arc. Trajectory 355 differs from trajectory 350 in that in trajectory 355 the camera moves further away from its focus point (and the user's current position) as it moves up the arc.
[0157] As the camera moves along one of the arcs, the camera rotates to maintain the point in front of the position indicator at the camera's focus. In some embodiments, the user can turn off the three-dimensional view and adopt a purely two-dimensional view. For example, some embodiments of the application allow the three-dimensional mode to be turned on and off through the use of a 3D button 360. The 3D button 360 is very useful for a turn-by-turn navigation feature, where the 3D button has the function of both an indicator and a two-state element. When 3D is turned off, the camera will maintain a 2D navigation experience, but when 3D is turned on, some top-down perspective may still be present when the 3D viewing angle is meaningless (e.g., when walking around a corner that would be blocked in 3D mode).
[0158] 2. User adjustment of camera height
[0159] In addition to (or instead of) having the navigation application control the camera (e.g., switching from 3D to 2D when walking around a corner), some embodiments allow the user to adjust the camera's level. Some embodiments allow the user to make command gestures using two fingers to adjust the camera's distance (height) and angle. Some embodiments even allow multiple types of gestures to control the camera. Figure 4 The diagram shows adjusting the distance of the virtual camera using pinch and expand gestures. The diagram is shown in three stages. In stage 401, the application displays a basic scene 410 with a virtual camera 412 at a default horizontal height for 3D viewing and a screen view 414 rendered from scene 410. The basic scene contains two buildings and a T-junction. In stage 401, the buildings are viewed from a 45-degree downward angle and at a specific height that makes them appear to be a specific size. A position indicator 416 is also shown at a specific size.
[0160] In stage 402, the user makes a gesture by placing two fingertips 420 close to each other on the screen of the device (on the screen view 424) and moving the fingertips apart while the fingertips are on the screen. Moving the fingertips 420 apart has the effect of making the map (both the portion between the fingers and the remaining portion of the map) larger. To make things in the map appear larger, the application zooms in on the virtual camera 412. In some embodiments, the line 450 along which the mapping application moves the virtual camera 412 is a line formed by the front of the virtual camera 412 and the focus point of the virtual camera 412. The mapping application of some embodiments moves the virtual camera 412 along a line formed by the front of the virtual camera 412 and a location in the 3D map 410 based on user input that zooms into (or out of) the view of the 3D map 410.
[0161] After zooming in for stage 402, the user decides to zoom out for stage 403. In this stage, the user has placed two fingers 430 on the screen and brought the two fingers closer together. Bringing the fingers closer together has the effect of zooming out the map (both the portion between the fingers and the remaining portion of the map). The zoom out adjustment is achieved by moving the virtual camera 412 farther away from the 3D map 410 along line 455. In some embodiments, the line 455 along which the mapping application moves the virtual camera 412 is a line formed by the front of the virtual camera 412 and the focus point of the virtual camera 412. The mapping application of some embodiments moves the virtual camera 412 along a line formed by the front of the virtual camera 412 and a position in the 3D map 410 based on user input that zooms into (zooms out of) the view of the 3D map 410.
[0162] Rendering a 3D map view using virtual camera 412 at this location results in 3D map view 434 in which buildings and roads appear farther away than the location shown in 3D map view 424. As shown by the dashed version of virtual camera 412, virtual camera 412 moves farther away from 3D map 410 along line 455.
[0163] In addition to zooming in and out controls, some apps allow users to change the angle of the virtual camera. Figure 5 An embodiment of a camera whose angle can be adjusted by gestures is shown. The diagram is shown in three stages 501-503. In stage 501, the camera is pointed downward at a 45-degree angle toward a scene 510. Scene 510 includes two buildings and a T-junction, shown in screen view 514. The buildings are shown from a specific angle and at a specific size. Location indicator 516 is also shown at a specific size.
[0164] In stage 502, the user has placed two fingers 520 approximately horizontally relative to each other on the screen and dragged them upward. This has the apparent effect of dragging the scene upward along with the fingers. The scene is raised by lowering the virtual camera 512 and changing its viewing angle from 45 degrees to 30 degrees. In the screen view 524, the building and location indicator appear taller than in stage 501.
[0165] After the user dragged the scene upward in stage 502, the user then drags the scene downward in stage 503. To do this, the user again places two fingers 530 on the screen and drags downward. This causes the scene to be dragged downward along with the fingers 530. The scene's descent is achieved by raising the virtual camera 512 and changing its angle downward relative to the scene 510 to 60 degrees. In stage 503, the camera 512 has moved further upward and is angled downward more than in stage 501. As a result, the buildings and location identifiers 536 appear even shorter and smaller in stage 503 than in stage 501.
[0166] In some embodiments, the mapping application provides inertia effects for different operations (e.g., panning, rotating, and transitioning from 2D to 3D). When a user provides a specific type of input (e.g., input that ends at a rate greater than a threshold rate) to pan a 3D map, the mapping application generates an inertia effect that causes the 3D map to continue panning and then slow down to a stop. In some embodiments, the inertia effect provides the user with a more realistic interaction with the 3D map that mimics real-world behavior.
[0167] Some embodiments allow the distance and angle of the camera to be controlled independently. For example, it allows the distance to be controlled by pinching and spreading the finger gesture, and allows the angle to be controlled by dragging a finger placed horizontally. Other embodiments use either gesture to set the distance or angle of the camera, with the other variable set automatically. While Figure 4 and Figure 5 Gestures performed in certain directions are shown to result in certain results, but in some embodiments, one or both of these gestures can be reversed. For example, in some embodiments, dragging a horizontally positioned finger downward can lower the camera rather than the scene. This will have the effect of moving the scene downward when the finger is moved upward, and the effect of moving the scene upward when the finger is moved downward.
[0168] II. Server-side generation of map tiles
[0169] In order to display both immersive and non-immersive 3D map presentations, some embodiments must generate various tiles for the client device to render to generate roads, buildings, and surrounding scenery. In some embodiments, examples of such tiles include road and building tiles for non-immersive 3D presentations, and navigation and building tiles for immersive 3D presentations.
[0170] Before generating these tiles, a set of servers must generate descriptions of the roads, buildings, and other geometric structures placed in each tile. This task involves multiple subtasks, such as (1) receiving map data from various vendors, (2) processing such data to produce one-dimensional (1D) roads, (3) smoothing the 1D road map, (4) defining data to specify junction points, (5) generating 2D road geometry and land cover, (6) smoothing the 2D road geometry, (7) generating data about buildings (e.g., estimated height data), (8) using such data to define building geometry, (9) constructing road geometry details (such as islands, lane markings, and distances between road geometry and land cover), and (10) identifying geometry edge node features and propagating such features.
[0171] The mapping service of some embodiments generates downloadable map tile data through offline processing of map data (e.g., data received from a map vendor). In some embodiments, the offline processing obtains map object location input (e.g., latitude / longitude data for roads, administrative boundaries, natural boundaries, etc.) and generates aggregated roads and relationships between the aggregated roads. Based on the aggregated roads and their relationships, the mapping service processing generates road geometry. The mapping service also uses the map object location input to generate the geometry of land covers (e.g., parks, oceans, states, etc.). Some embodiments use scalable distributed processing to create downloadable map tiles based on geometric vector data. Those of ordinary skill in the art will recognize that the "offline" processing described in this application can be performed by a mapping service computing device that is actually connected to the network through which the mapping application requests tile data, but is used to indicate that the processing is not performed in response to a user request for a tile.
[0172] Figure 6 The mapping service process 600 of some embodiments is conceptually illustrated. As mentioned, some embodiments perform this process offline to generate various map tiles, which can then be sent to client devices in response to real-time requests. The process can be performed periodically (e.g., hourly, daily, weekly) or can be performed any time new data becomes available (e.g., through updates received from providers of map data, corrections received from users, etc.).
[0173] The mapping service process 600 includes a 1-D road processor 605, a road geometry generator 610, a land cover geometry generator 615, a building geometry generator 617, and a tile generator 620. One of ordinary skill in the art will recognize that Figure 6The various modular operations shown in can all be executed on a single server, or can be distributed across several computing devices. In fact, some operations (e.g., tile generation) can be performed as distributed processing operations across multiple computing devices.
[0174] In addition to processing operations, Figure 6 Also shown are road segment data 625, land cover data 630, and building data 633. Road segment data 625 stores data defining the location and characteristics of roads. In some embodiments, each road segment is assigned one or more names (e.g., "Lincoln Blvd.", "CA-1"), location data indicating the path of the road segment, and attributes of the road segment (e.g., speed limit, width, number of lanes, road type, etc.). In some embodiments, the location of junctions (also referred to as junctions) is part of the road segment data. Land cover data 630 stores data defining the location and characteristics of various types of non-road land cover. Land cover may include bodies of water (e.g., rivers, oceans, lakes, swimming pools, etc.), administrative entities (e.g., state, country, city, park, etc. boundaries), regional designations (e.g., rural / urban / suburban, desert / mountain / forest, etc.), or other data describing the land between roads. Building data 633 in some embodiments stores the location of buildings and data related to the buildings. For example, building data may include surface elevation data and ground elevation, from which the building height can be calculated.
[0175] Road segment data 625, land cover data 630, and building data 635 may be received from a single source (e.g., a single map data provider) or from several sources (e.g., several different providers). For example, a first provider may provide road data for a first region, while a second provider provides road data for a second region, a third provider provides land cover data for both the first and second regions, and a fourth provider provides elevation data (or other combinations, such as multiple different providers of land cover data for different types of information).
[0176] The 1-D road processor 605 receives the road segment data 625 and generates one or more road graphs from the data. In some embodiments, the road graph links a set of road segments with road segment junctions. In order to generate a road graph, the 1-D road processor 605 identifies the road segments that should be combined into aggregate roads (i.e., based on the name, starting point and end point and the common attributes of the road segments). For example, when the end of a first road segment shares a location with the beginning of a second road segment, the road segments have at least one common name, and the road segments share at least some attributes, then the road processor 605 combines the road segments into aggregate roads. When the data required for additional processing (e.g., elevation data, speed limit data, number of lanes) is missing in the road segment data, the 1-D road processor additionally fills in the data.
[0177] In addition to forming aggregated roads, the 1-D road processor 605 identifies and generates angle information for the junctions of the roads (as described below) and identifies other relationships between the roads (e.g., merging dual carriageways, establishing overlapping relationships). The connected set of junctions and road segments forms a road graph.
[0178] The road geometry generator 610 generates geometry for roads (i.e., a set of vertices used to draw roads at a client device) based on a road graph using various processes. In some embodiments, the road geometry generator 610 grows road geometry from road centerlines (e.g., using lane number, width, and offset data) and then performs various operations to refine the geometry. As examples, the road geometry generator 610 smoothes transitions between road segments with different numbers of lanes, creates more realistic rounded corners at intersections, smooths road merges to remove redundant road blocks and gaps between roads, and removes overlaps between disjoint independent roads. Additionally, in some embodiments, the road geometry generator 610 assigns specific features to polygon vertices and / or edges, such as marking tunnel entrances, marking interior road edges (e.g., between road segments), marking sidewalks, and marking bridge frames within which shadows should be drawn.
[0179] In some embodiments, the land cover geometry generator 615 uses both the land cover data 630 and the road segment data 625 to create the geometry for the land cover (i.e., the set of vertices defining the polygons). After growing the geometry for the land cover, some embodiments use various operations to resolve the intersections between the geometries. For example, when combining data from different sources, the position data indicating the intersection of objects may not be perfectly aligned and therefore there may be gaps between the object geometries or overlaps of the geometries. Some embodiments use different operations to resolve the intersections between different geometries, depending on the type of object. In addition, the land cover geometry generator 615 uses the road segment data to fill in the land cover geometry and ensure that no gaps are left between the land cover and the road. Some embodiments cause the land cover geometry to grow toward the road beyond its marked intersection, stopping the geometry at the road centerline location. Although this creates an overlap between the land cover geometry and the road geometry, in some embodiments, the client mapping application includes instructions for rendering the road geometry on top of the land cover geometry.
[0180] The building geometry generator 617 of some embodiments uses the building data 633 to generate the building geometry. In some embodiments, as mentioned, in addition to the location of the building, the building data 633 also includes ground elevation and ground elevation data. To generate the building geometry, some embodiments calculate the building height for each point within the building's location. The building geometry generator 617 retrieves the ground elevation and subtracts it from the ground elevation to calculate the building height. In other embodiments, the building geometry generator 617 (or another module) uses 3D satellite data to calculate the height data. To calculate the height of the building as a whole, the building geometry generator 617 calculates the total height as the average of the individual calculated heights at different points, plus a deviation factor constant multiplied by the standard deviation of the point heights. The deviation factor is a constant determined based on ground truth (e.g., data determined at an actual location) and experimentation.
[0181] Some embodiments also determine whether the building is flat or non-flat (e.g., has a pointed roof). When the standard deviation of the point heights is above a threshold (which may also be based on ground truth and experimentation), the building geometry generator 617 designates the roof as non-flat. When the standard deviation of the point heights is below the threshold, the building geometry generator 617 designates the roof as flat. When generating geometry vertices, some embodiments create pointed roofs (e.g., triangular prisms or pyramids) for non-flat buildings.
[0182] Road, land cover, and building geometry are sent to a tile generator 620. In some embodiments, the tile generator 620 creates several tiles for the map area at different levels of detail (i.e., zoom levels). Some embodiments define tile location boundaries for different zoom levels (e.g., where a tile at a first zoom level is contained within four tiles at the next zoom level) and then use distributed processing techniques to assign different geometries (both roads and land cover) to the various tiles. After assigning the geometries to the tiles (each geometry can be assigned to one or more tiles at each zoom level), the tile generator 620 uses additional distributed processing to generate and compress the tiles. In some embodiments, map tiles contain vector data that describes polygons generated to render the data into a 2D or 3D map. To reduce the amount of vector data (and thus reduce the file size for easier transmission), some embodiments use a transient rasterization process that reduces the vector data to raster information and then re-vectorizes the data using fewer vertices.
[0183] As shown, the tile generator 620 outputs tile data 635 to a storage device. The stored tile data 635 is data accessed by the client mapping application so that the application generates a map for viewing by the user. As shown, the tile data of some embodiments includes non-immersive map tiles 640, immersive map tiles 645, building tiles 650, and traffic tiles 655. The non-immersive map tiles 640 and the immersive map tiles 645 provide different levels of data. In some embodiments, the mapping application includes an immersive 3D view and a non-immersive 3D view, wherein the immersive 3D view provides additional details (e.g., an asphalt view of the road, more realistic land cover and buildings, a closer simulation of the driving field of view, etc.). Some embodiments use independent tile sets for different views, wherein the immersive map tiles 645 include a greater level of detail about the road and land cover. In addition to road segment and land cover data, map tiles can also contain additional data. For example, tiles can include various sets of label data (e.g., road labels, location labels, land cover labels, etc.). As shown, in some embodiments, the tile data 635 also includes building tiles indicating the geometry used to draw buildings (based on height and location information), and traffic tiles that are regularly updated with traffic information 655. Some embodiments use building tiles for both immersive and non-immersive 3D views.
[0184] A. Road Data and Road Map Generation
[0185] As described above, some embodiments receive road segment data (ie, from one or more sources) and generate a road map based on the road segments. Figure 7A process 700 of some embodiments for generating a road map is conceptually illustrated. Figure 8 To describe process 700, Figure 8 Various data structures of some embodiments for roads within a map area are shown. These data structures, which will be described in more detail below, include road segments 800, edges 805, roads 810, junctions 815, and a road map 820.
[0186] As shown, process 700 begins by receiving (at 710) a set of road segments and junctions for a map area. In some embodiments, process 700 (or a similar road map generation process) is run separately for different map areas (e.g., for states, rectangular geographic areas, land masses, etc.). Road segments for a map area can be received from a single source or from several different sources. In some embodiments, a road segment is a consistent stretch of road with a single consistent set of attributes (i.e., the same number of lanes, speed limits, etc.). When a road attribute changes (e.g., the speed limit changes from 45 mph to 35 mph, or a lane is added), a new road segment is defined.
[0187] The join point for the map area is typically received from the same source as the roads that meet at the join point. In some embodiments, the join point defines an intersection of at least two roads—that is, two or more road segments not only pass through the same location (which can be determined based on the road segment data), but the road segments actually intersect with each other so that a vehicle can transition from one road segment to the other at the join point.
[0188] Figure 8 The data structure 800 of some embodiments of a road segment and the data structure 815 of a junction are shown. As shown, a road segment includes a segment ID (i.e., a unique identifier), one or more names, geometry information, and attribute information. The geometry information (which is different from the road geometry created to define vector data) defines the path and other geometric information about the road segment. As shown, the geometry information includes centerline path data (e.g., a sorted string of coordinates defining the center of the road), start and end junction information, parameters for indicating width and offset relative to the centerline, and a function capable of evaluating both sides of the road at any point along the road segment. In some embodiments, this is a function on the road segment class that uses the centerline, offset, and width information to calculate the position of both sides of the road. Although the figure shows road drawing data including start and end junctions, some embodiments do not define one junction as the start and one junction as the end, but instead only indicate two junction IDs as the end points (or only indicate one junction ID as the end point if the road segment reaches a dead end).
[0189] Attribute information describes metadata about the road segment, such as the road type (or functional road class, which defines the importance level of the road, from highway down to the proposed route), the number of lanes, the speed limit, the relative elevation of the road (which may contain references to one or more other road segments and / or junctions, indicating whether the current road segment extends below or above the referenced object), the height of the road (related to the identification elevation), the form of the road (which defines the route as a dual carriageway, single carriageway, walkway, steps, connecting road, turnout, etc.), restrictions (for example, long-distance call restrictions, vehicle type restrictions, indication that the road is private, etc.).
[0190] In addition, if Figure 8 As shown in , some embodiments define edges 805 for each road segment. The edge data structure contains a reference (e.g., a pointer) to the road segment to which the edge corresponds, an orientation flag indicating whether the edge is oriented in the same direction as the road segment, and functionality for calculating the right and left sides of the edge using the width and centerline data of the referenced road segment. In some embodiments, this functionality exists as a function on the edge class. Edges are, in some embodiments, data constructs used to create road graphs. For road segments with both a start and end junction, some embodiments define two edges (one in each direction) so that each junction can reference an edge leading from the junction.
[0191] like Figure 8 As shown in , the junction data structure 815 includes a unique junction ID, an ordered list of edges with associated angles, a location, and an elevation level. While this data structure illustrates a list of edges, some embodiments additionally include a list of road segments referenced by segment IDs, from which the edge list can be generated. In some embodiments, the junction ID, the list of road segment IDs, and the elevation levels are received from a road data source, while the angles and edge list are calculated by a mapping service. Because each road segment is associated with one or more edges, the mapping service process can determine the appropriate guide edge (the edge that is directed away from the junction) to use for the junction list. For many surface street junctions, relative elevations are not particularly meaningful, as the junction and all roads at that location will be at the same level. However, at intersections such as interchanges, or when road segments pass under or over highways, elevation levels indicate which junctions are on top of other junctions and / or road segments.
[0192] After receiving the road segments, the process defines (at 710) a road map for the map area based on the road segments and junctions. Figure 8As shown in , the road graph data structure 820 includes a set of road segments and a set of junctions. In some embodiments, the application traverses the edges and junctions to identify a connected set of edges, and then maps these edges to road segments. Other embodiments use the start and end junctions stored in each of the road segments, and a list of road segments stored for each junction, to define the connectivity of road segments and junctions. Some embodiments define a single road graph for a map area. However, when there are multiple disjoint connected sets, some embodiments will define multiple road graphs.
[0193] Next, process 700 performs (at 715) pre-processing on the road map to align the road segments at the junctions. In order to properly identify the angles of the roads at the junctions, and to generate polygonal geometry for the roads, the roads that intersect at the junctions should all end / start at exactly the same location. However, the road centerlines for each of the road segments that meet at a particular junction may not end at exactly the same location in the received map data. Therefore, for each junction, some embodiments calculate the average location where the road segment paths end and assign that location to the junction at its location. Additionally, the mapping service process modifies the road segment path data for each of the road segments that meet at the junction so that the centerlines all end at the calculated average location.
[0194] Figure 9 Shown is the example of this type of modification for two road segments 905 and 910 that meet at a junction in two stages 920 and 930. As shown in the figure, each road segment includes a start junction and an end junction, wherein the end junction of the first road segment 905 and the start junction of the second road segment 910 are identical (both with reference to the junction with ID 16). However, as shown in the first stage 920, the centerline paths of the first road segment 905 and the second road segment 910 do not start / end at the same place (and, in fact, do not intersect at all). Therefore, some embodiments identify the average position of the end that should meet. For a junction with only two road segments, this position will be midway along the line between the two path endpoints, as shown by position 915 in the figure. Each of the end vertices of the centerline has a pair of coordinate values (e.g., (x, y) coordinate values or geographic location coordinates (latitude, longitude)). Some embodiments calculate an average value between all the end vertices of each coordinate, and use this average value as the average position of the junction.
[0195] After identifying the junction location, the mapping service processes the modified road segment so that the center lines all end at the identified location. The second stage 930 shows an example of such a result. In some embodiments, when the road segment is received, the centerline path may not have vertices at fixed distances. The mapping service processes of some embodiments standardize the road segment vertices so that each vertex is at a fixed distance (e.g., 10 meters), thereby achieving a smoother (less tortuous) road polygon. Some embodiments identify a vertex that is a fixed distance (e.g., 50 meters, 100 meters, etc.) from the end for a particular road segment and keep that point fixed while modifying the other vertices between that point and the end vertex. In some embodiments, the vertices are moved less as the vertex becomes further away from the end point. That is, a vector is calculated for the movement of the end vertex to the identified location, and the other vertices are moved by fewer iterations of that vector.
[0196] With the pre-processing complete, the process combines (at 720) the road segments into aggregated roads. Some embodiments determine when the road segments are actually different parts of the same road so that the road segments can be treated as a single road for the purposes of defining maps (e.g., for generating labels) and for generating certain aspects of polygons for the roads. In some embodiments, the process uses various attributes and geometry data of the road segments to determine whether two road segments should be combined.
[0197] Some embodiments traverse a road graph to combine road segments into aggregated roads. When traversing the graph and entering a junction from a particular road segment, a mapping service process determines which of the other road segments is the best match to be a continuation of the current road segment. Some embodiments score each of the road segments and aggregate roads in a greedy manner.
[0198] To calculate the score for a particular road segment, the mapping service process evaluates the angular difference between the incoming road segment and the particular road segment, the compatibility of the road segment names, and compares road attributes. For angle, the closer the road segment is to 180° (i.e., a continuation of a straight line through a junction), the higher the score the segment will generally score; typically, a road will continue in a straight or approximately straight line through a junction without turning. Some embodiments perform a fuzzy comparison between the road names associated with each road segment. That is, the process uses an approximate string matching technique to compare the string of the incoming road segment with the string of the particular road segment. In some cases, one or more of the road segments may have multiple names, and the process performs a comparison of each to find one or more best matches, and uses these best matches in calculating the score (e.g., using "San Diego Freeway South" to compare with "San Diego Freeway South" instead of with "1-405 South").
[0199] In addition, in some embodiments, at least some of the road attributes are compared to calculate a comparison score. For example, the mapping service processing of some embodiments compares the road type (i.e., highway, main road, secondary road, etc.), the number of lanes, the speed limit, the form of the road (i.e., one-way road, two-way road, etc.). Once the scores are calculated, some embodiments select the segment with the highest score and determine whether the highest score is above a threshold for continuing the road. In addition, some embodiments identify the best segment selected and perform a comparison between the selected segment and each of the other segments. The process aggregates the roads only when the first segment is the best match for the second segment and the second segment is the best match for the first segment. This prevents an incoming segment that actually ends at a "T" intersection from joining one of the segments that actually continues through the intersection.
[0200] Figure 10 An example of three road segments 1005, 1010, and 1015 that are candidates for converging into a road at a junction 1020 is shown. As shown, the first road segment 1005 has the name "Welsh Avenue," a speed limit of 40 mph, and three lanes, and is a main road. The second road segment 1010 also has the name "Welsh Avenue" and three lanes, and is also a main road, but has a speed limit of 35 mph. The third road segment 1015 has the name "Saltair Street," a speed limit of 35 mph, and two lanes, and is characterized as a secondary road. If the initial entry road segment is the third road segment 1015, then the third road segment 1010 will be the higher-scoring road segment of the two possible continuations. However, for the second road segment 1010, the third road segment 1015 will not be the best road continuation because the first road segment 1005 will have a much higher continuation score (the angle, road name, road type, and number of lanes are identical, and the only difference is the 5 mph speed limit). Thus, the best result at this junction 1020 is for a road in which the third road segment 1015 is part of what will terminate, and the first road segment 1005 and the second road segment 1010 are joined together as a single road. Other embodiments do not perform an interaction test, but instead set a sufficiently high threshold that a bad match is unlikely. Furthermore, when traversing the road graph, some embodiments allow a road segment to be moved from one aggregated road to another when a better match occurs later in the traversal (e.g., if road segment 1015 is found to continue into road segment 1010, but then passes through road segment 1005 again to reach junction 1020, then road segment 1010 can be moved from the first road to the new road that also contains road segment 1005).
[0201] Figure 8The road data structure 810 of some embodiments is shown. As shown, the road data structure 810 includes an ordered list of edges (by which a road object contains references to the constituent road segments of the road object). In addition, the road structure includes functions for evaluating the road centerline and both sides of the road at any point along the road. In some embodiments, the centerline function exists as a function of the road class that references the centerline information of the constituent road segments (and can also interpolate between any slightly mismatched centerlines). The roadside function also exists as a function of the road class that uses the roadside function from the segment objects contained by the road object.
[0202] return Figure 7 After combining the road segments, process 700 derives (at 725) additional road description data. In some embodiments, this involves filling in missing data in the road segment information. Because some embodiments implement road segments (and other data) as objects, in some cases the object variables will have null values because the data has not yet been provided by the map data provider. For example, speed limit data, road width, number of lanes, or other parameters may be missing. Some embodiments automatically fill in the values of at least some of the missing data for a road segment using other information from the road segment or information from adjacent road segments, where possible. For example, some embodiments may fill in missing speed limit information for a road segment based on the speed limits of adjacent road segments (e.g., if a first road segment has a speed limit of 55 mph and a third road segment has a speed limit of 35 mph, the middle road segment may be filled in with a speed limit of 45 mph).
[0203] Other embodiments only populate the data necessary to generate the road geometry, such as the number of lanes and road width information. For example, some embodiments may use the adjacent road segments within the aggregated road to generate the number of lanes (e.g., if the road segments on both sides of a particular road segment have a specific number of lanes, then that specific number of lanes may also be assigned to that particular road segment). For road width, some embodiments use the number of lanes (if any) to assign a width to the road (e.g., assuming each lane is 4 meters wide). On the other hand, some embodiments assign road width based on road type (i.e., highways have a first width, major roads have a second width, etc.). In fact, some embodiments derive the number of lanes from the road type (e.g., highways are always assigned three lanes, etc.), and then generate the width based on the number of lanes.
[0204] In addition, some embodiments fill in missing data for height or elevation information. As described above, in some cases the map data specifically indicates that a particular road segment passes over (or under) another road segment, or passes over (or under) a junction, or that a particular junction passes over (or under) another junction. However, this relative elevation data may be missing in some cases. When a first road segment (or junction) and a second road segment (or junction) have centerline paths that intersect in a flat plane, but no junction is defined for the roads, then the paths must be at different elevations. When absolute elevation data is included from a map data provider, this data can be used to determine relative elevation data (i.e., defining a relationship where a first road segment having a lower absolute elevation passes under a second road segment having a higher absolute elevation).
[0205] When relative elevation data cannot be determined for a particular set of road segments that intersect in the plane but do not meet at a junction, some embodiments create a random relative ordering for the road segments. In other embodiments, the possible relative orderings can be derived from the road types of the non-intersecting roads, but attributes of other nearby road segments can also be used. Figure 11 A scenario is shown where a first road segment 1105 ("I-405N") intersects (in plan) with a second road segment 1110 ("Welshire Avenue") and a third road segment 1115 ("Main Street"). However, in this case, the map data does not indicate any junction between these road segments. The I-405N road segment has an elevation of 1 and the Wilshire road segment has an elevation of 0. However, the Main Street road segment has a null elevation (i.e., this information is not provided). Based on the fact that the I-405N road segment travels above the Wilshire Avenue road segment, some embodiments also assume that I-405N will travel above Main Street. As such, the mapping service operation automatically populates the elevation of the road segment data structure for Main Street with a value of 0, or with a relative order indicating that the Main Street road segment travels below the I-405N road segment.
[0206] In addition, process 700 establishes (at 730) relationships between roads and road segments. In some embodiments, establishing relationships includes identifying semantic data, such as the opposite sides of a dual carriageway, and defining connections between related roads. In some embodiments, the process identifies roads that contain road segments with the same or similar names (e.g., "I-5N" and "I-5S"), roads that are marked as dual carriageways (i.e., in the form of road segments), and roads that are within a threshold distance of each other and within a threshold angle of parallelism. Figure 12Two road segments 1205 and 1210 are shown, running parallel in opposite directions, a distance X apart, and both named "Main Street." Assuming the distance X is less than a threshold distance for roads with the same name, these roads will be linked as a connected dual carriageway, and the turn between the roads can be classified as a U-turn rather than an independent left turn. Additional semantic data that can be propagated includes assigning names to connecting roads / turnouts (e.g., freeway entrance, freeway interchange, right turnout, etc.).
[0207] Finally, process 700 generates (at 735) a smoothed version of the road. In some embodiments, the centerline data may be noisy. For example, a five-mile stretch of highway may be straight in reality, but the data may have slight repetitions here and there. In addition, because the road is defined as a line between centerline vertices, bends may appear where the road actually has smooth turns. Furthermore, when road segments join at a junction, this can create sharp turns that should actually be smoother. Because this noise is undesirable, some embodiments apply a smoothing operation controlled by the road characteristics to the roads in the road map (e.g., one road segment at a time, or to converging roads). To perform the smoothing, a variety of different operations may be performed by different embodiments. Some embodiments use a smoothing operation (e.g., Laplace smoothing) that moves each vertex in the road a distance based on the position of its neighboring vertices.
[0208] Some embodiments allow more smoothing for faster roads because it is expected that these roads (e.g., highways) will be smoother and less likely to have sharp corners. Some embodiments use speed limit data to control smoothing. For example, some embodiments allow different deviations of the centerline from the received vertex data for roads with different speed limits (e.g., allowing up to 5 meters of deviation for a road with a speed limit of 65, and 2 meters of deviation for a road with a speed limit of 30 mph). Figure 13 A winding road 1310 is shown with a speed limit of 25 mph. The result of applying a smoothing operation to the road is shown as a modified road 1320. On the other hand, Figure 14 A similar curvy road 1410 is shown with a speed limit of 60 mph. The result of applying a similar smoothing operation to the road is shown as modified road 1420. As shown, modified road 1420 shows a higher level of curve smoothing because the speed limit of the road is greater. Some embodiments use other characteristics to modify the level of smoothing allowed, such as the road type (which can serve as a proxy for the speed limit).
[0209] Additionally, some embodiments limit smoothing in special cases. For example, a road might make a 90° turn (e.g., where two roads end at a junction in a populated area and join together). In such cases, some embodiments fix two points close to the junction (one on each side of the 90° turn) and smooth the small curve between these two points, rather than smoothing the road too much.
[0210] B. Generate road geometry
[0211] After generating a road map from road segments and junctions, aggregating roads in the road map, filling in missing data, smoothing aggregated roads, and the like, offline processing by the mapping service of some embodiments generates geometry for the roads in the road map. In some embodiments, the geometry includes a set of vertices that define the edges of polygons to be rendered by a client mapping application. These geometries are included in vector map tiles that are sent to a device running the mapping application so that the mapping application can render the resulting polygons. The mapping service processing is initially based on the geometry of the road segment definitions (i.e., centerline vertices and width data), but also performs various conditioning operations on the geometry, such as smoothing polygons to create more realistic rendered roads and intersections, eliminating overlaps, and annotating vertices with data that marks features to be drawn (e.g., shadows, bridge frames, curbs, sidewalks, tunnels, etc.).
[0212] Figure 15 Conceptually illustrated is a process 1500 of some embodiments for generating road geometry to be used to generate road polygons on a client mapping application. As with process 700, various embodiments perform process 1500 as an offline process that can be executed on a single computing device or distributed across several computing devices.
[0213] As shown, process 1500 begins by receiving (at 1505) a road map for a map area. Figure 8 As shown, a road map is a connected set of road segments and junctions. In some embodiments, process 700 (or similar road map generation processes) is run separately for different map regions (e.g., for states, rectangular geographic regions, continental blocks, etc.).
[0214] The geometry generation process 1500 generates (at 1510) initial polygonal representations of road segments in the road graph. Some embodiments unfold the segmented linear centerline representation (which should be reasonably smoothed in a process such as process 700 described above after inserting vertices at fixed intervals and then smoothing them) into polygons that follow the path of a road with a width specified by its road segment data structure. The resulting polygons are essentially structures with parallel sides that may appear curved. However, these polygons will need to be adjusted in many cases to generate more realistic road animations.
[0215] In some embodiments, polygon adjustments include smoothing polygons to create more realistically rendered roads and junctions, eliminating overlaps, and annotating vertices with data marking features to be drawn (e.g., shadows, bridge frames, curbs, tunnels, etc.). The next several operations in process 1500 describe these adjustment operations in a particular order. However, one of ordinary skill in the art will recognize that once the initial polygons are grown, these operations can be performed in a variety of different orders.
[0216] Process 1500 adjusts (at 1515) the geometry at the joins. Generally speaking, smoothing is primarily desired at the joins of the road graph, in addition to overlaps between two or more geometries. Within a road segment, the smoothing of the vertices described above should prevent any kind of imposed artifacts or irregularities. Some embodiments traverse the road graph and, at each join, determine whether any modifications to the geometry are necessary to ensure a smooth, attractive rendered road at the client mapping application (and then make those modifications). These joins may simply involve roads that continue in a straight line where the speed limit changes (if the road width remains constant, this will not involve any smoothing in some embodiments), but may also include freeway merges, roads that intersect at various angles, intersections between large and small roads (of which freeway merges are one example), lane extensions, and the like.
[0217] The following will be accompanied by some examples of Figure 1 Various types of these join smoothing examples are described below. Some embodiments perform specific smoothing operations when a road expands or contracts its number of lanes. In some embodiments, the mapping service process first determines whether the join consists of only two edges, and then determines whether the two edges are connected within the road. If so, and the number of lanes changes, some embodiments perform one of the following transition smoothing operations to prevent unnatural-looking sharp corners.
[0218] Figure 16The first operation of smoothing lane extension junctions is shown as being performed over three stages 1610-1630, wherein one of the two sides of the road remains straight (i.e., "aligned"). As shown in the first stage 1610, in this example, a first road segment 1605 connects to a second road segment 1615. The first road segment 1605 has data indicating that the first road segment has two lanes, while the second road segment 1615 has data indicating that the second road segment has three lanes. In some cases, this data is indicated in the road segment data received from the map data provider. When the map data provider includes lane information, in some cases the provider also includes lane connectivity data. In some cases, the lanes are numbered (e.g., from left to right), and the connectivity data indicates which lane in the first road segment (e.g., road segment 1605) continues into which lane in the second road segment (e.g., road segment 1615). As shown in the first stage 1610, in this case, lane 1 of the first road segment 1605 continues into lane 1 of the second road segment 1615, while lane 2 of the first road segment continues into both lane 2 and lane 3 of the second road segment. This indicates that the lane extension occurs on the right side of the road (the "right" based on the direction of travel of the road). In various embodiments, this lane connectivity data can be stored with the junction point, or with one or both of the road segments.
[0219] The second stage 1620 shows the initial growth of the segment geometry 1625 and 1635 for the first and second segments. As can be seen, the second segment geometry 1635 is wider than the first segment geometry 1625, and all of this extra width is on the right side of the road, while the left side of the road is aligned. When provided with lane connectivity data, some embodiments generate geometry with extra lanes appearing in the correct locations. However, when rendered using this geometry, the asphalt will appear to be at a 90-degree angle, and the lanes will appear inexplicably. While highway lanes do occasionally begin this way, typically highway lanes expand smoothly from existing lanes.
[0220] Therefore, some embodiments automatically taper lane expansion, as shown in the third stage 1630. Some embodiments identify a point along the edge of the narrower road segment geometry (the geometry with fewer lanes) and taper from that point to the beginning of the wider road segment edge. Some embodiments use a fixed distance to identify this point (e.g., 25 meters, 50 meters, etc.). The taper can be a straight line (as shown) or a curve. Additionally, some embodiments use a transition within the wider geometry, or have the transition spread across both geometries.
[0221] Figure 17The result 1700 of smooth lane expansion as rendered by the client-side mapping application of some embodiments is shown (e.g., on a display of a portable device). For this figure, and many of the figures within this section, the result for a particular road segment as rendered on the client-side mapping application is shown. In these figures, various controls and features that would typically be present on the mapping application display (e.g., a user location indicator (or navigation locator), route lines, additional land cover and building features, UI controls, etc.) are not shown in order to highlight various aspects of the road segment. Figure 17 In the example, the asphalt makes a smooth transition from two lanes to three lanes, and the lane markings begin a little before the lanes are fully formed. As described below, some embodiments add road boxes (e.g., curbs, guardrails, etc.) and lane markings. In various embodiments, instructions for drawing the road boxes can be transmitted to the client mapping application as additional geometry or as annotations to the geometry, where the instructions for rendering are based on the annotations stored on the client device.
[0222] Figure 18 The second operation is shown as being performed over three phases 1810-1830 to smooth the lane extension junction, where the road is extended on both sides. As shown in the first phase 1810, in this example, the first road segment 1805 connects to the second road segment 1815. In this case, there is no lane connectivity data, but the first road segment 1805 has two lanes, while the second road segment 1815 has three lanes. The lane count data can be provided directly from the map provider or derived from the provided road width or road type.
[0223] The second stage 1820 shows the initial growth of the segment geometry 1825 and 1835 for the first and second segments. As can be seen, the second segment geometry 1835 is wider than the first segment geometry 1825, and the extra width is split between the two sides of the road. When no lane connectivity data is available, some embodiments use a default lane extension that splits the difference between the two sides of the road. However, when rendered using this geometry, the asphalt will appear to have a 90-degree angle on both sides. While highway lanes may occasionally begin this way, typically highway lanes expand smoothly from existing lanes.
[0224] Thus, some embodiments automatically taper the lane extension on both sides, as shown in the third stage 1830. Some embodiments identify a point along the edge of the narrower road segment geometry for each side of the road and taper from that point to the beginning of the wider road segment edge. Some embodiments use a fixed distance to identify this point (e.g., 25 meters, 50 meters, etc.). The taper can be a straight line (as shown) or a curve. Additionally, some embodiments use a transition within the wider geometry, or have the transition spread across both geometries.
[0225] Figure 19 The result of the smooth lane extension as rendered by the client mapping application 1900 (e.g., on a display of a portable device) is shown. The asphalt makes a smooth transition from two lanes to three lanes. Some embodiments bend two of the lanes in one direction and add a new lane on the other side, so that in some embodiments, the lane markings remain continuous (e.g., a random edge is picked for the new lane, or a default value is used).
[0226] The above example is a situation where the junction of interest has only two road segments. For junctions with three or more road segments, some embodiments identify special cases and for other cases step through the list of road segments in the junction and perform corrections on the geometry for subsequent road pairs. One such special case occurs when a two-way single-lane road is split into a pair of corresponding dual carriageways. This often occurs when a central divider is introduced into a dual lane road (one lane in either direction). Some embodiments gradually taper the road outward (from a single lane to a dual carriageway) at the outer edge and identify a merging point for the inner edge of the two dual carriageways. Some embodiments use lane information to determine the merging point (for example, if one of the dual carriageways is one lane and the other is two lanes), while other embodiments use a default value of a point at the center of the single lane (at its end).
[0227] When stepping through a list of road segments at a junction, some embodiments perform different corrections based on the angle between the two road segments. Some embodiments select the first road segment at the junction, then correct the corner formed by the left side of the road segment geometry (as viewed from the junction), where the right side of the road segment geometry (as viewed from the junction) for the second road segment is directly counterclockwise relative to the first road segment. The second road segment is then set as the first road segment, and the process is repeated for the next corner until all corners have been corrected.
[0228] The following will be accompanied by some examples of Figure 1 The various corrections performed for various junction angles are described below. These figures, although showing junctions with several segments, also show the geometry for only two segments that coincide with each other in the example.
[0229] Figure 20 The diagram shows a process of performing a gradual reduction operation to smooth the corners between the road segment geometries over three stages 2010-2030 when the angle between the road segments is greater than a first threshold angle (e.g., 170°). In such cases, some embodiments automatically reduce the edge of one of the roads to meet the edge of the other road. As shown in the first stage 2010, in this example, the first road segment 2005 intersects the second road segment 2015 at a junction point and intersects two additional road segments 2025 and 2035.
[0230] The second stage 2020 shows the initially grown segment geometries 2040 and 2045 for two road segments 2005 and 2015. The geometries of the other two road segments are not shown in this figure because the mapping service processing of some embodiments corrects one corner of a junction at a time, and this example shows the correction between road segment 2005 and road segment 2015. The mapping service processing of some embodiments will subsequently perform corrections for the corners between road segment 2015 and road segment 2025, the corners between road segment 2025 and road segment 2035, and finally the corners between road segment 2035 and road segment 2005. As shown at this second stage 2020, the initial corner between geometry 2040 and geometry 2045 involves a visually jarring (and unrealistic) discontinuity, in part because the roads have different widths (although some embodiments perform the same process for roads of the same width).
[0231] Thus, some embodiments automatically taper the edge of the narrower segment to the edge of the wider segment, as shown by the modified geometry 2040 in the third stage 2030. In this case, because the wider segment is the more counter-clockwise of the two segments, the processing tapers the left side of the narrower segment to the right side of the wider segment (for this and subsequent examples, the "right" and "left" sides of a particular segment are as viewed from the perspective of a vehicle exiting the junction along the particular segment). When the wider segment is the more clockwise of the two segments, the processing tapers the right side of the narrower segment to the left side of the wider segment. To perform the taper operation, some embodiments fix a point along the edge of the narrower segment geometry and taper from that point to the beginning of the edge of the wider geometry. Some embodiments use a fixed distance to identify that point (e.g., 25 meters, 50 meters, etc.). The taper can be a straight line (as shown) or a curve. This operation is similar to Figure 16 The taper shown in and can also be performed when only two roads of different widths meet at a junction.
[0232] Figure 21Result 2100 of a smooth junction as rendered by a client-side mapping application of some embodiments is shown (e.g., on a display of a portable device). The asphalt indicates a smooth right turn from a first road segment (shown at the bottom of the display) to a second, wider road segment (shown at the top right of the display). In addition, the display includes asphalt (and lane markings, road boxes, etc.) for the other two road segments forming the intersection. In various embodiments, instructions for drawing the road boxes can be transmitted to the client-side mapping application as additional geometry or as annotations to the geometry, where the instructions for rendering are based on the annotations stored at the client device.
[0233] Figure 22 The diagram shows that, over three stages 2210-2230, a projection operation is performed to smooth corners between road segment geometries when the angle between the road segments is less than a first threshold angle and greater than a second threshold angle (e.g., between 120° and 170°). In such cases, some embodiments automatically project the edge of one of the roads to meet the edge of the other road. As shown in the first stage 2210, in this example, the first road segment 2205 intersects the second road segment 2215 at a junction point and intersects two additional road segments 2225 and 2235.
[0234] The second stage 2220 shows the initially grown segment geometries 2240 and 2245 for two road segments 2205 and 2215. The geometries of the other two road segments are not shown in this figure because the mapping service processing of some embodiments corrects one corner of a junction at a time, and this example shows the correction between road segment 2205 and road segment 2215. The mapping service processing of some embodiments will subsequently perform corrections for the corner between road segment 2215 and road segment 2225, the corner between road segment 2225 and road segment 2235, and finally the corner between road segment 2235 and road segment 2205. As shown at this second stage 2220, the initial corner between geometry 2240 and geometry 2245 involves a visually jarring (and unrealistic) discontinuity, in part because the roads have different widths (although some embodiments perform the same process for roads of the same width).
[0235] In view of the Figure 20, some embodiments of the mapping service processing cause the narrower road to taper to the wider road, in which example, the mapping service processing causes the edge of the wider road to be projected onto the narrower road, as shown by the modified geometry 2240 in the third stage 2230. In this case, because the wider road segment is the more counterclockwise of the two road segments, the processing causes the right side of the geometry 2245 to be projected onto the centerline (or left side, in some embodiments) of the geometry 2240. When the wider road segment is the more clockwise of the two road segments, the processing causes the left side of the wider road segment to be projected onto the centerline (or right side, in some embodiments) of the narrower road segment. Given that the taper operation involves changing the angle of the edge of one of the road segments, in some embodiments the projection only extends the projected edge, as shown in stage 2230.
[0236] Figure 23 Result 2300 of a smooth junction as rendered by a client-side mapping application of some embodiments is shown (e.g., on a display of a portable device). The asphalt indicates a smooth right turn from a first road segment (shown at the bottom of the display) to a second, wider road segment (shown at the top right of the display). In addition, the display includes asphalt (and lane markings, road boxes, etc.) for the other two road segments forming the intersection. In various embodiments, instructions for drawing the road boxes can be transmitted to the client-side mapping application as additional geometry or as annotations to the geometry, where the instructions for rendering are based on the annotations stored at the client device.
[0237] Figure 24 The diagram shows a clipping operation performed over three stages 2410-2430 to eliminate redundant roads from the road segment geometry when the angle between the road segments is less than a first threshold angle and greater than a second threshold angle (e.g., between 30° and 80°). In such cases, some embodiments automatically clip the edges of one or both of the roads to avoid overtaking the other. As shown in the first stage 2410, in this example, the first road segment 2405 intersects the second road segment 2415 at a junction and intersects the additional road segment 2425.
[0238] The second stage 2420 shows the initial growth of the segment geometry 2440 and 2445 of the two segments 2405 and 2415. The geometry of the other segments is not shown in this figure because the mapping service of some embodiments processes the correction of one corner of the junction at a time, and this example shows the correction between the segment 2405 and the segment 2415. Figure 25A correction is shown between road segment 2415 and road segment 2425. As shown at this second stage 2420, the initial corner between geometries 2440 and 2445 is fine, but the right side of geometry 2445 exceeds the right side of geometry 2440, in part because geometry 2440 is narrower than geometry 2445.
[0239] Thus, the mapping service processing of some embodiments clips the left side of the more clockwise segment back to the centerline of the more counterclockwise of the two segments, and correspondingly clips the right side of the more counterclockwise segment back to the centerline of the more clockwise segment, as shown by the modified geometries 2440 and 2445 in the third stage 2430. In some embodiments, only the edge of the wider segment is clipped, but other embodiments clip both geometries as shown. As shown in the third stage 2430, this removes an extra "ear" of the road that would otherwise extend unrealistically beyond the intersection.
[0240] Figure 25 The intersection operation is shown to be performed over three stages 2510-2530 to smooth the corner between the road segment geometry 2445 and the road segment geometry 2505 of the road segment 2425. Some embodiments perform the intersection operation for angles greater than 180° due to gaps left between the road segment geometries. In such cases, some embodiments identify intersections between tangents to the road edges and extend or clip the road geometry so that the edge ends at the intersection.
[0241] The second stage 2520 shows that Figure 24 , and the initially grown segment geometry 2505 of segment 2425. As shown, this (1) leaves a gap between the two geometries, and (2) results in the wider segment 2445 extending beyond segment 2505. The second stage also shows a dashed line representing the intersection of the projection of the left side of the more clockwise geometry 2445 and the projection of the right side of the more counterclockwise geometry 2505.
[0242] Some embodiments clip or project the edges of the geometry to the intersection, as shown in the third stage 2530. For the narrower road segment 2505, the edges are extended to fill the gap, while for the wider road segment 2445, the edges are clipped to prevent the geometry 2505 from exceeding.
[0243] Figure 26 The result of the smooth join from the first two graphs as rendered by the client mapping application of some embodiments (e.g., on a display of a portable device) is shown 2600. The asphalt indicates a smooth turn between each of the subsequent roads, with no missing sections or extra pieces of asphalt.
[0244] Figure 27 The diagram shows a gradual reduction in angles between road segment geometries over three stages 2710-2730, which in some embodiments are performed to smooth the corners between road segment geometries in a situation that conforms to the characteristics of a highway on-ramp. In this example, a first road segment 2705 continues straight into a third road segment 2725 at a junction, where the second road segment 2715 enters the junction at a shallow angle relative to the first road segment 2705. This situation, in some embodiments, conforms to the profile of a highway merge: the first and third road segments are connected roads, are wider than the second road segment, and the angle between the first and second road segments is less than a threshold angle (e.g., 30°).
[0245] The second stage 2720 shows the initial growth of the segment geometry 2735-2745 for the three segments 2705-2725. The two highway segment geometries 2735 and 2745 are significantly wider than the on-ramp segment geometry 2740. In addition, the third segment geometry 2745 is wider than the first segment geometry 2745 to accommodate the merging lane. However, the initial geometry therefore leaves a gap between the on-ramp geometry 2740 and the third segment geometry 2745.
[0246] Therefore, the mapping service processing of some embodiments causes the outer side (left side) of the second geometric structure 2740 to taper to the right of the third geometric structure 2745, as shown in the third stage 2730. In this case, the taper is shown as a curved roadside, but different embodiments may use other methods (e.g., using straight lines to connect road segments). Some embodiments also use a taper from the left side of the first geometric structure 2735 to the right side of the third geometric structure 2745, as shown in the rendering result shown below.
[0247] Figure 28 The result 2800 of a smoothed highway merge junction as rendered by a client-side mapping application of some embodiments (e.g., on a display of a portable device) is shown. The asphalt indicates a smooth merge, rather than the choppy display that would be rendered as a result of the initial rectangular geometry. Furthermore, the figure shows that some embodiments render different roads differently. For example, some embodiments render highways in a different color (or hue) than other roads, including highway on-ramps.
[0248] In addition to modifying geometry using operations such as tapering, projection, and clipping, some embodiments use techniques to create rounder corners at certain types of joins. Rather than using techniques to modify already generated segment geometry to create rounded corners, some embodiments generate new geometry to create these rounded corners. In some embodiments, modifying already generated segment geometry can be extremely processing intensive.
[0249] Figure 29 The generation of this additional geometry at the junction is shown over three stages 2910-2930 to create more realistic rounded corners at the junction. The first stage 2910 shows a junction with four road segments 2905-2935 that intersect at right angles. The second stage 2920 shows the initial geometry for these road segments. In this case, all road segments have the same width, so the geometry all has the same shape and size. Because the angles between the road segments are all right angles, none of the projection, clipping, or tapering described above is required.
[0250] The third stage 2930 shows the generation of additional geometry that overlaps the road segment geometry and generates the rounded corners of the intersection. Specifically, the additional geometry 2940 (shown as a solid line, while the road segment geometry is represented by a dashed line at this stage) has edges radiating from the junction point locations toward points at a fixed distance from the 90° corner formed by the road segment geometry. The additional geometry 2940 also has curved lines between these points (or straight lines between enough vertices to approximate curved lines so that the intersection will be rendered with a more realistic curved appearance).
[0251] Figure 30 The result 3000 of a smoothed intersection as rendered by a client-side mapping application of some embodiments (e.g., on a display of a portable device) is shown. The asphalt does not make awkward 90° turns between road segments, but instead has a more curvy, pleasing appearance. In addition, the intersection includes stop lines and centerline markings, the generation of which is described below.
[0252] return Figure 15, process 1500 also eliminates (at 1520) road overlaps. In some cases, two (or more) initially grown road geometries will incorrectly overlap. The geometries of road segments that intersect at a junction will overlap, and when two road segments that do not form a junction but span the same latitude / longitude coordinates (e.g., roads at stacked levels, roads passing above / below each other, etc.), their geometries will overlap (as projected onto a flat plane representing the Earth's surface). However, when most other geometries overlap, this is likely due to incorrect data and should be corrected. Without correction, the rendered map will indicate to the user of the mapping application that the two roads intersect when they do not actually intersect.
[0253] The most common case of such overlap is when two corresponding dual carriageways overlap (e.g., northbound and southbound or eastbound and westbound highways, opposite sides of an arterial road divided by a median, etc.). In most cases, the geometry of such corresponding dual carriageways should be very close to each other and not actually overlap. However, because the road segment geometry is based on the centerline and width data of the road segment, if either of the information is slightly inaccurate, the geometry may not line up correctly. For example, when neither a lane count nor a width is provided for a road, some embodiments derive the width based on the road type. However, if a road is actually narrower than is typical for its road type, the road geometry may end up being larger than the actual road in the real world. Other situations that may cause such overlap include a highway and its branches, a road that ends at a highway rather than passing above or below the highway, etc.
[0254] Different embodiments use different techniques to remove these overlaps in road segment geometry. For example, some embodiments push the two roads apart at the location of the overlap, while other embodiments reduce the width of the roads at the overlap. Additionally, after moving the roads apart, some embodiments add a center divider geometry between the road geometry when the road is a corresponding dual carriageway.
[0255] Figure 31The diagram shows two roads being pushed apart as they overlap over three stages 3110-3130. The first stage 3110 shows a first road segment 3105 and a second road segment 3115, each as a sequence of vertices. As described above, some embodiments use a constant distance between the centerline vertices of a road in order to standardize operations performed using vertices. When the road segment data is received by a mapping service from a data provider, the vertex data representing the centerline of the road segment may be defined at varying distances (e.g., with gaps of 10 meters, 25 meters, 50 meters, etc.) in some cases. Some embodiments automatically define vertices at predetermined intervals (e.g., every 10 meters) when ingesting the road segment data. This ensures that (i) the road will have a more curved, realistic appearance than if the vertices were spread out, because the mapping service of some embodiments handles defining straight edges between vertices. In addition, it makes the above-mentioned smoothing operation more successful by providing more points to move during the operation. In addition, defining vertices at fixed intervals enables more useful tests for road overlap.
[0256] Stage 1 3110 also shows the roadsides generated for the road geometries of segments 3105 and 3115. As can be clearly seen, the two generated road geometries overlap for a portion. To detect such overlap, some embodiments generate a capsule around each of the edges of the road centerline (the line between subsequent vertices) and test the capsules for overlap. Stage 1 3110 shows four such capsules: capsules 3125 and 3135 for the first segment 3105, and capsules 3140 and 3145 for the second segment 3115. In some embodiments, a capsule is defined to encompass the edge between two subsequent vertices and extend to the width of the road (using the centerline and width data of the segment). Although shown as an oval (i.e., ellipse) in this example, different embodiments use different shapes for the capsules, such as a rectangle or other parallelogram, other ellipsoids, etc.
[0257] As shown at the first stage, the capsule 3125 of the first edge of the first segment 3105 overlaps with the capsule 3140 of the first edge of the second segment 3115. Additionally, the capsule 3135 of the second edge of the first segment 3105 does not overlap with the capsule 3145 of the second edge of the second segment 3115. Additionally, although not shown in this figure, between those edges having capsules shown, the capsules of the two edges of each segment will likely also overlap. For simplicity in illustrating the modifications to the segments, Figure 31
[0258] These additional overlaps are not shown or illustrated in . In some embodiments, when an overlap is detected between the capsules of two road segments, the mapping service process applies a "force" to the vertices surrounded by the capsules in order to push the road segments away from each other. Some embodiments use a spring-mass equation to determine this force, where a larger overlap between capsules results in a larger outward force. That is, the greater the overlap, the more the conceptual spring mass "compresses," resulting in a larger outward force. Some embodiments apply this force to each of the four vertices within the two overlapping capsules, and then gradually reduce the force to zero as it reaches nearby vertices. The second stage 3120 shows the application of these forces, with the largest force arrows applied to the two vertices inside each of the capsules 3125 and 3140, and smaller force arrows applied to the vertices on either side.
[0259] In some embodiments, mapping services processes each set of identification overlapping capsules and identifies the force applied to each vertex due to that single overlap. After identifying the forces on different vertices, mapping services processes the forces integrated on each vertex to derive the new centerline path for the road segment. Processing then performs the same capsule test to determine whether any overlap between the road segments still exists. If there is still overlap, mapping services processes additional force applied to the vertex and again pushes the vertex apart. The third stage 3130 shows that two road segments 3105 and 3115 have been pushed apart so that the capsule around the edge of the first road segment does not overlap with the capsule around the edge of the second road segment at all. As shown, the roadside of the two road segment geometries is now independent and does not overlap at all.
[0260] Other embodiments use different mechanisms to move vertices apart. For example, even in the case of multiple overlapping capsules, some embodiments move vertices as soon as the first overlap is detected, and then move to the next set of vertices once the first identified set has been pushed far enough apart to avoid any overlap. In addition, while the examples show that force is applied to vertices near those edges of the overlap, some embodiments only move vertices at the edges that actually cause the overlap (i.e., only vertices within capsules 3125 and 3140). Furthermore, different embodiments may use different equations (i.e., other than spring-mass equations) to determine how far to push vertices away from each other, and use different equations to calculate how far to move surrounding vertices.
[0261] Moving the centerline vertices of the road segments is one option used by some embodiments to handle overlapping road geometries.On the other hand, some embodiments reduce the width of the road near the overlap rather than moving the centerline (ie, reducing the size of the road geometry).
[0262] Figure 32The width of two road segments is reduced as they overlap over three stages 3210-3230. The first stage 3210 shows a first road segment 3205 and a second road segment 3215, each as a sequence of vertices (similar to Figure 31 ). As described above, some embodiments use a constant distance between centerline vertices of a road in order to standardize operations performed using vertices. When road segment data is received by a mapping service from a data provider, the vertex data representing the centerline of the road segment may be defined at varying distances (e.g., with intervals of 10 meters, 25 meters, 50 meters, etc.) in some cases. Some embodiments automatically define vertices at predetermined intervals (e.g., every 10 meters) when ingesting the road segment data.
[0263] The first stage 3210 additionally shows the roadsides generated for the road geometries of segments 3205 and 3215. As can be clearly seen, the two generated road geometries overlap for a portion. To detect such overlap, some embodiments generate a capsule around each of the edges of the road centerline (the line between subsequent vertices) and test the capsules for overlap. The first stage 3210 shows two such capsules: capsule 3225 for the first segment 3205 and capsule 3235 for the second segment 3215. The capsule is defined in some embodiments to enclose the edge between two subsequent vertices and extend to the width of the road (using the centerline and width data of the segments). Although shown as an oval (i.e., ellipse) in this example, different embodiments use different shapes for the capsule, such as a rectangle or other parallelogram, other ellipsoid, etc. Some embodiments may use Figure 31 The solution shown in and Figure 32 Both solutions shown in were tested, and the same capsule was used for each solution.
[0264] As shown at the first stage, the capsule 3225 of the edge of the first segment 3205 overlaps with the capsule 3235 of the edge of the second segment 3215. In addition, although not shown in this figure, below those edges with overlapping capsules shown, the capsules of the two edges of each segment are likely to overlap as well. For simplicity in showing the modifications to the segments, Figure 32
[0265] These additional overlaps are not shown or described in the . In some embodiments, when an overlap is detected between the capsules of two road segments, the mapping service process reduces the width of the road segment at the location of the overlapping capsule. As shown in the second stage 3220, some embodiments move the roadside at the location of the overlap inward toward the centerline a specific distance. Some embodiments move that portion of the roadside of a particular road segment a distance proportional to the amount of overlap. The distance can be just enough to reduce the overlap at the location, or that amount plus an additional margin so that there is no gap between the two road segment geometries. In addition, some embodiments will only cause the overlapping roadside to move inward, while other embodiments also symmetrically modify the other side of each road (so that the centerline remains in the center of the road).
[0266] To prevent the rendered road on the client mapping application from appearing to have abrupt large blocks removed from the road, some embodiments select a point along the roadside in either direction (e.g., 100 meters, 500 meters, etc. from the affected vertex) and keep these points fixed. The mapping service process then tapers the width reduction from the overlapping portion to the fixed point.
[0267] The third stage 3230 shows the road geometry of road segments 3205 and 3215 after the width of the road has been reduced. As shown, the road geometry no longer overlaps at all (at least in the portion shown in the figure). Some embodiments, after detecting an overlap between any two pods, perform a process to reduce the width at that location and cause the width reduction to taper off. The process then retests that edge pair and continues to identify whether any overlapping road segments still exist. On the other hand, other embodiments first identify all overlapping segments, reduce the width at each such segment, and calculate the different tapers required for each width reduction.
[0268] In addition to the geometry corrections performed at 1515 and 1520 to smooth the joins and eliminate road overlap, process 1500 also generates (at 1525) road markings for display on the road. These road markings may include lane markings, median strips, stop lines, and / or crosswalks at intersections, etc., in some embodiments.
[0269] For example, some embodiments automatically add a median strip between two lanes of traffic. As described in the previous section, some embodiments use segment data to identify road segments (or aggregated roads) as corresponding sides of a road (e.g., opposite directions of a highway or sides of an arterial road separated by a median strip, etc.). In some embodiments, the segment data indicates that the road is a two-lane road (sometimes called a divided highway, as opposed to a single lane that carries traffic in both directions), and other characteristics (both geometric data and attribute data of the road segments) indicate the likelihood that the two roads are corresponding two-lane roads.
[0270] Such dual carriageway roads typically have two sides separated by a median of various types. For example, different areas may use a concrete median, a grass median, a painted median (e.g., different combinations of yellow markings), a divider, etc. Some embodiments generate geometry that fills the gap between the corresponding dual carriageway roads and then assigns these geometric properties so that the geometric properties are rendered as the median. Some embodiments use the locations of the vertices of the roadside to determine the locations of the median vertices so that the median is precisely aligned with the edges of the road.
[0271] Different embodiments render medians differently. For example, some embodiments use a single median type for all dual carriageways, while other embodiments use different types. Much like roads, land cover, etc., a style sheet stored at the client can define different median styles that are drawn in different ways. Median geometry from the mapping service is labeled with different styles corresponding to the different ways the map is rendered by the client mapping application (e.g., arterial road medians, freeway medians, etc.).
[0272] Figure 33 and Figure 34 A dual carriageway is shown with two different types of central dividers. Figure 33 Two rendered roads 3305 and 3315 corresponding to dual carriageways are shown. In addition, the application displays a design of yellow markings (specifically, a pair of double yellow lines with a diagonal line connecting the yellow markings) representing a central dividing strip 3310 between the two roads. Figure 34 A concrete median strip 3405 drawn between two dual carriageway carriageways 3305 and 3315 is shown.
[0273] In addition to medians, some embodiments generate geometry for various types of road markings (e.g., lane dividers, stop bars, etc.). In some embodiments, this includes lane markings as shown in the rendering results described above. To generate lane markings for a road segment, some embodiments use lane count information stored in the road segment data structure (which may be derived from width data or road type data). Additionally, dedicated lanes, such as carpool lanes, may be indicated in the road segment data and may have generated geometry.
[0274] Additionally, some embodiments identify junctions where the driver will likely have to stop at a stop sign or stop light. The mapping service of some embodiments receives this data from the map provider (e.g., as information stored in the junction data). Thus, a particular junction may indicate a 4-way stop, a 2-way stop (singling out a particular road segment as having a stop sign), a traffic light, etc. In some embodiments, the mapping service processing derives this information based on the type of road at the junction. For example, when a connecting road intersects a main line road, some embodiments assume that the connecting road has a stop sign and that the main line road has a clear right of way. When two main line roads intersect, the mapping service processing assumes that the intersection will be controlled by a traffic light and adds stop line markings to all road segments at the junction.
[0275] Figure 35 Shown is the example of the geometry of this type of junction 3500. When all road section geometries of intersection all have the same width and are arranged at 90 ° of angles, it is easy (as shown below) to generate stop line so. However, four road sections 3505-3520 are not aligned so well. On the contrary, road sections have different widths, and road section 3510 intersects with junctions at different (non-right angle) angles. In such cases, for each specific road section, the mapping service processes the line closest to the intersection perpendicular to the center line of the specific road section, and contacts the two sides of the specific road section geometry when it does not intersect with other road section geometries. Although shown as for more complicated junctions, in simpler cases, some embodiments also use this process to identify stop line position.
[0276] Figure 35 The dashed lines 3525 in the figure show these lines at the junction 3500. The mapping service process then generates geometry for stop lines that end at these identified lines (dashed lines 3525). For one-way road segments (which have two-way travel), the process only generates stop lines that extend halfway across the segment. In addition, some embodiments push the stop lines back a specific distance (e.g., two feet, five feet, etc.) from the edge of the intersection (in this figure, the edge is represented by the dashed line).
[0277] Figure 36 The result 3600 of the join point 3500 as rendered by the client mapping application of some embodiments (e.g., on a display of a portable device) is shown. As shown, the thick white line is drawn in Figure 35 The position that the dotted line of 1 is identified is drawn half across each road section. The inner (towards the intersection) edge of the stop bar geometry is drawn at the indicated line, wherein the stop bar is extended all the way into the road section (away from the intersection). Some embodiments also push the stop bar away from the intersection by a fixed distance. In addition, lane markings are generated so that they stop at the stop bar or not far before the stop bar.
[0278] Return again Figure 15 , the road geometry creation process 1500 next specifies (at 1530) annotation information about the road geometry and then ends. In some embodiments, this annotation information is stored on polygon vertices (or edges, as the two are essentially interchangeable), which guides the mesh building module and / or rendering engine of the client mapping application to apply specific features to the rendered output. For example, some embodiments specify different types of road boxes (i.e., objects located on both sides of the road) through the road geometry vertices, instructing the client application to draw curbs, sidewalks, bridge boxes (and corresponding shadows), tunnels, overlay roads, etc. In some embodiments, the instructions for drawing these boxes are stored on the client device as part of the mapping application. The client mapping application then reads the information stored on the specific vertex and generates the appropriate road box for the edge corresponding to the vertex.
[0279] Shown below are several examples of different annotations for some embodiments of road vertices that are used to indicate the type of box to be drawn. Figure 37 shows an example of using vertex annotations to specify the bridge frame of an interchange where one road runs over another and the roads do not form a junction. Figure 37 As shown, two road geometries 3705 and 3710 intersect in flat space. However, the segment data specifies that the segment for which geometry 3705 was generated has a higher elevation (using either absolute elevation data or relative elevation data) than the segment for which geometry 3710 was created. Furthermore, while one segment is shown for each of the two roads, in some embodiments, several segments will be used to define the road (e.g., for road 3705, the first segment goes from an elevation of 0 to an elevation of 1, the second segment has an elevation of 1, and the third segment goes from an elevation of 1 to an elevation of 0).
[0280] Figure 37 Also shown are several vertices 3715-3735 that define the sides of road geometry 3705 and 3710. Each particular vertex defines a leading edge from that particular vertex to the next vertex in sequence. Thus, for example, the annotation for vertex 3735 defines how the client mapping application will draw the edge from vertex 3735 to vertex 3730, the annotation for vertex 3730 defines how the client mapping application will draw the edge from vertex 3730 to vertex 3725, and so on. Different vertices along a single road segment can have different annotations, so that portions of a single road segment will be drawn with different features (e.g., different road boxes).
[0281] In this case, vertex 3715 (as well as the other vertices of the road segment geometry 3710) specifies that it is a vertex of a connecting road and specifies that a default box is used. In some embodiments, different types of roads (e.g., connecting roads, arterial roads, highways, etc.) may not only be drawn differently (e.g., different colors / textures for asphalt, different widths, etc.), but may also have different boxes (e.g., curbs for connecting roads). While this illustrates the actual designation of a default box, some embodiments do not have a specific annotation on the vertex for a default box (or any other characteristic of the vertex that provides an annotation for it).
[0282] Vertices 3725-3735 are shown as specifically designating bridge frames for their respective edges of geometry 3705, while vertex 3720 designates a default arterial road frame. Just as different types of roads can have different default frames, different types of roads can also have different bridge frames. That is, a client-side mapping application can include instructions for drawing different interchanges for arterial roads, highways, and so on. Vertex 3720, although located at the edge of the interchange area, does not specify a bridge frame, but rather a default road frame for its edge. This is because its edge, for that side of road geometry 3705, points away from the interchange toward the next vertex to the left (not shown).
[0283] Figure 38 Result 3800 is shown (e.g., on a display of a portable device) of road segments 3705 and 3710 as rendered by the client-side mapping application of some embodiments. In this case, the asphalt of the connecting road is rendered using the same color / texture as the asphalt of the main road, but some embodiments may render the roads differently. In addition, a lighter-colored curb is drawn along the edges of both roads for locations where vertices are assigned default road boxes. However, for the overpass portions, the client application draws the overpass bridges according to some embodiments. In addition, the application renders shadows on the asphalt of the connecting road to show the overpasses more realistically. If the user were to rotate the view of the map for these roads, similar bridge boxes would be rendered for the other sides of the overpasses. In addition, while shadows appear only for the overpasses of the connecting road, some embodiments may also render shadows below the main road as it moves from the ground up to the overpass.
[0284] Figure 39 An example of using vertex annotations to designate edges of road segment geometry as inner edges is shown, indicating that a box should not be drawn for the guide edge from the annotated vertex. Some embodiments use such annotations to prevent client mapping applications from drawing curbs in the middle of roads where roads meet at junctions. Figure 39As shown, two road segment geometries 3905 and 3910 meet at a junction. In this case, the road segments are connected as continuous portions of the aggregated road; however, in some embodiments, the same road box annotation can also be used at a junction that represents an intersection of more than one road.
[0285] Figure 39 Also shown are several vertices 3915-3925 that define the sides of road geometry 3905. Other road geometry 3910 is also defined by vertices, including vertices located at the same point as vertices 3915 and 3920. As in the previous figure, each specific vertex defines a leading edge from that particular vertex to the next vertex in sequence. Thus, for example, the annotation for vertex 3915 defines how the client mapping application will draw the edge from vertex 3915 to vertex 3920, the annotation for vertex 3920 defines how the client mapping application will draw the edge from vertex 3920 to vertex 3925, and so on.
[0286] In this case, all vertices 3915-3925 (as well as the other vertices of the road segment geometry 3910) specify that they are vertices that connect to the road. In addition, vertices 3925 and 3920 specify that a default box is used. While this shows the actual designation of a default box, some embodiments do not have a specific annotation on the vertex for a default box (or any other characteristics of the vertex that provide an annotation for it). In addition, vertex 3915 specifically indicates that an inner box is used (i.e., the guide edge associated with vertex 3915 is an inner edge of the road). In some embodiments, the inner box annotation instructs the client mapping application to not draw any road box at all for these edges.
[0287] Vertex 3920, although located at a corner near the inner edge of the road geometry, is assigned a default box rather than an inner box. This is because its edge is oriented toward vertex 3925. Furthermore, in some embodiments, the vertex of road geometry 3910 will be located at the same point as vertices 3915 and 3920, with the vertex at the same location as vertex 3920 being assigned an inner box (assuming the direction of the edge is the same for geometry 3910 as for 3905).
[0288] Figure 40Result 4000 is shown (e.g., on a display of a portable device) of road segments 3905 and 3910 as rendered by the client-side mapping application of some embodiments. In this case, the asphalt of the first segment blends smoothly into the asphalt of the next segment, with no visual indication of the transition from one segment to the next. The application also draws the curb as a default box along either side of the road. In addition to the polygons for the lighter-colored curb, some embodiments also draw details, such as a small shadow generated by the curb, to provide a more three-dimensional feel to the display, at least in some modes of the client-side mapping application. Although the inner (i.e., facing other segments) edge of the road is designated so that the curb and other road boxes are not drawn in the middle of the road, some embodiments also use inner annotations for roads that do not have boxes on their outer edges, such as rural roads that do not have curbs or other boxes that are at the same level as the grass near the road.
[0289] Figure 41 An example of using vertex annotations to specify tunnels on a road segment is shown, indicating that a tunnel should be drawn on the road segment. Some embodiments use such annotations to indicate the entrance / exit of a tunnel, as well as the middle of a tunnel, so that the client mapping application will draw the tunnel when rendering the road segment. Figure 41 As shown, two road segment geometries 4105 and 4110 meet at a junction. In this case, the road segments are connected as continuous portions of the aggregated road; however, in some embodiments, the same road box annotation can also be used at a junction representing an intersection of more than one road.
[0290] Figure 41 Also shown are several vertices 4115-4130 that are used to define both sides of the road geometry 4110. Other road geometry 4105 is also defined by vertices, including vertices located at the same point as vertices 4115 and 4130. In some embodiments, these vertices are annotated to specify a default box or inner box for different vertices. As in the previous figures, each specific vertex defines a guiding edge from the specific vertex to the next vertex in sequence. Thus, for example, the annotation for vertex 4115 defines how the client mapping application will draw the edge from vertex 4115 to vertex 4120, the annotation for vertex 4125 defines how the client mapping application will draw the edge from vertex 4125 to vertex 4130, and so on.
[0291] In this case, all vertices 4115-4130 (as well as the vertices of the road segment geometry 4105) specify that they are part of the main road. In addition, vertex 4130 specifies the use of a tunnel entrance box, while vertices 4115-4125 specify the use of a tunnel box. In some embodiments, the tunnel entrance box annotation specifies that certain edges (edges of the road segment geometry 4110 that cross the road surface) are the entrances to the tunnel and should be drawn as such. The tunnel box annotation of some embodiments specifies that the edges are the sides of the tunnel and therefore the tunnel should be drawn across the road at the locations of these edges.
[0292] Figure 42 Result 4200 is shown (e.g., on a portable device's display) of road segments 4105 and 4110 as rendered by a client-side mapping application of some embodiments. In this case, the first road segment is displayed with asphalt and a curb, which is the default rendering of some embodiments. However, for the second road segment, the application draws a curved tunnel entrance and then a circular overlay of the road segment. At the other end of the road segment, the application renders the tunnel exit. In addition, the application renders a shadow below the entrance, causing the asphalt to appear darker at that location.
[0293] Figure 43 An example of a sidewalk on one side of a road segment is shown using vertex annotations. The figure shows a single road geometry 4305 for a road segment that runs in a straight line. As shown, a number of vertices 4310-4325 are used to define the two sides of the road geometry 4305. Each specific vertex defines a leading edge from the specific vertex to the next vertex in sequence. Thus, for example, the annotation for vertex 4310 defines how the client mapping application will draw the edge from vertex 4310 to vertex 4315, the annotation for vertex 4320 defines how the client mapping application will draw the edge from vertex 4320 to vertex 4325, and so on. Different vertices along a single road segment can have different annotations so that portions of a single road segment will be drawn with different features (e.g., different road boxes).
[0294] In this case, vertices 4310 and 4315 specify that they are vertices of a connecting road and specify that a default box be used. As described above, different types of roads may have different default boxes (e.g., curbs for connecting or arterial roads, guardrails for certain other types of roads, etc.). While this illustrates the actual designation of a default box, some embodiments do not have specific annotations on the vertices for a default box (or any other characteristics of the vertex that provide annotations for it).
[0295] Vertices 4320 and 4325 specifically indicate sidewalk road boxes for respective edges of geometry 4305. Just as different types of roads can have different default boxes, different types of roads can also have different sidewalks. That is, the client mapping application can include instructions for drawing different sidewalks for connecting roads, arterial roads, and so on (e.g., wider sidewalks for arterial roads).
[0296] Figure 44 Result 4400 is shown for road segment 4305 as rendered by a client-side mapping application of some embodiments (e.g., on a display of a portable device). In this case, the asphalt of the main road is rendered in the same manner as shown above, with only the curb drawn on the left side (with the default road box). However, for the right side of the road, in addition to the curb, the application also renders the sidewalk. In addition to the polygon for the lighter-colored curb, some embodiments also draw details, such as a small shadow generated by the curb, to provide a more three-dimensional feel to the display, at least in some modes of the client-side mapping application.
[0297] The above examples illustrate different types of annotations that can be applied to road segment geometry. In some embodiments, multiple types of road boxes can be applied to a single edge by annotating the vertices of the edge with several types of annotations. Figure 43 In the above example, the sidewalk annotation implies that the client application should render both the curb and the sidewalk.
[0298] Figure 45 An example of using several annotations for a single vertex of some embodiments is shown. Specifically, Figure 45 An example of using vertex annotations to specify an overpass that includes a pedestrian walkway on one side of the bridge is shown. Geometries 4505 and 4510 are similar to Figure 37 The geometric structures 3705 and 3710, the latter (connecting road) passes under the former (arterial road). In fact, Figure 37 The difference is that vertices 4515-4525 include annotations specifying both the bridge frame and the sidewalk frame.
[0299] Figure 46 Result 4600 shows road segments 4505 and 4510 as rendered by a client mapping application of some embodiments (eg, on a display of a portable device). Figure 45Similarly, the asphalt of the connecting road is rendered using the same color / texture as the asphalt of the main road, and a lighter-colored curb is drawn along the edges of both roads. For the overpass section, the client application renders not only the overpass and its shadow, but also the sidewalk on top of the overpass, on the side of the road annotated with a sidewalk. In this case, the other side of road segment 3705 is not annotated with a designated sidewalk, so the application does not render the sidewalk.
[0300] In the previous example, the mapping service only specified the box type for the vertex. In some embodiments, the mapping service can also specify various variables for different types of boxes to modify the rendering of the box. For example, the height of the bridge box can be specified in the annotation data (i.e., based on the road segment elevation information). In addition, sidewalk width, curb height, guardrail height, tunnel thickness, curb color, etc. can vary in different embodiments.
[0301] Figure 47 An example of using vertex annotation data to change the properties of a road box for a road geometry 4705 is shown. Specifically, the road geometry 4705 includes two vertices 4710 and 4715 on one side and two vertices 4720 and 4725 on the other side. The two vertices 4710 and 4715 specify a rendered sidewalk box for the client application and include a width of 1 meter for the sidewalk. On the other side of the road, the two vertices 4720 and 4725 also specify a rendered sidewalk box for the client application, this time with a width of 1.5 meters.
[0302] Figure 48 Result 4800 of road segment 4705 as rendered by a client mapping application of some embodiments (e.g., on a display of a portable device) is shown. Figure 44 However, the rendered road now has sidewalks on both sides, and the sidewalks are different widths (as specified in the annotation data).
[0303] The above examples illustrate the use of vertex annotation data to specify road box data for some embodiments. However, because vertices are used in some embodiments to describe other map objects (e.g., buildings, land cover polygons, etc.) in map tiles, these other objects may also include vertex annotations in their descriptions. For example, such annotations may be used to specify building details (e.g., different color faces, different types of moldings, etc.), or to specify aspects of land cover (e.g., how to draw different boundaries of land cover polygons).
[0304] After the annotation information is specified, the road geometry creation process ends. As mentioned above, in some embodiments, the specific order of operations for creating the road geometry may be different. In addition, some embodiments may use parallel processing to perform some or all road geometry operations on several road segments (or sets of road segments) at once.
[0305] C. Server-side Generation of Intersection Data
[0306] Having described server operations that directly derive road geometry (e.g., defining aggregate roads and creating road geometry), we will now discuss the generation of intersection data using the junction data mentioned above. Upon requesting a route, the client mapping application, in some embodiments, receives data related to each intersection traversed by the route, indicating the different branches of the intersection and noting which branches the route traverses to enter and exit the intersection. In some embodiments, this intersection information is stored by a mapping service, which the mapping application accesses to retrieve map data as well as route and navigation information.
[0307] As described above, the mapping service of some embodiments receives data for each junction of a specified road segment. In some embodiments, the mapping service automatically generates additional data for each of these junctions stored in the map data, and in some embodiments, converts the junction data into intersection data. This junction information simplifies each received junction (or a collection of related junctions that are considered a single intersection in the real world) into a collection of branches that leave the junction at different angles. When a user requests a route through a mapping application operating on a device, the device sends the request to the mapping service, which generates a route from the starting location to the ending location. In some embodiments, the mapping service also uses the intersection data to generate turn-by-turn navigation instructions for the route. The mapping service identifies the intersection to be maneuvered and modifies the intersection data specific to the maneuver performed during the route. This data is then sent to the user device on which the client mapping application is running. The following sections first introduce the creation of navigation data for the route, then discuss the pre-calculation of intersection data performed by the mapping service (so that the data can be used to generate navigation data), and finally describe the specific types of modifications made to the intersection data for the requested route.
[0308] 1. Navigation data creation
[0309] Figure 49 Conceptually illustrates the operations performed by a mapping service of some embodiments to generate a route for a requesting device and provide the route along with navigation instructions to the requesting device. Figure 49This operation is shown proceeding through six stages 4910-4960, in which the mapping service receives a request for a route, generates a route, generates intersection data for the route, and provides the route to a device, which uses this information to display navigation instructions.
[0310] As shown in the figure, Figure 49 Each stage of the process depicts a device 4905 and a mapping service 4900. In some embodiments, the device 4905 may be a handheld device (e.g., a smartphone, a tablet device, etc.), or may be a dedicated navigation device (e.g., a built-in car navigation system, a portable navigation device, etc.). In addition, in some embodiments, the device 4905 may be a non-portable device, such as a desktop computer or other non-portable computing device.
[0311] Mapping service 4900 is a service to which device 4905 connects (e.g., via a wired connection, a wireless connection such as a cellular network, Wi-Fi, etc.) to request and receive map data, route data, turn-by-turn navigation data, and additional information (e.g., information about places located on a map, etc.). As shown, mapping service 4900 stores map data 4915 and intersection data 4925, and includes, among other modules (not shown), a map generator 4935 and a route generator 4945.
[0312] Map data 4915 provides the data from which viewable map areas and routes are generated. This map data, in some embodiments, includes latitude and longitude data, name data, and descriptive data related to roads and other pathways (e.g., walkways, ferry routes, bike paths, etc.), natural features (e.g., rivers, lakes, mountains, etc.), points of interest (e.g., buildings, businesses, parks, etc.), and other map features. In some embodiments, for example, a pathway is defined as a series of latitude / longitude vertices, names, and descriptive data. This descriptive data may include the type of road (e.g., whether the pathway is a single-carriageway or part of a dual-carriageway, whether the pathway is a one-way path), the road category to which the pathway belongs (e.g., motorway, local road, private road, bike path, etc.), and other information. In some embodiments, this map data is compiled and provided to the mapping service by an external source (i.e., a map provider), while in other embodiments, the mapping service provides its own map data. Map data may also be a mix of externally provided data and internally generated data. Furthermore, map data may include geometric data for various map constructs, such as roads, land cover, and the like.
[0313] Intersection data 4925 provides pre-tabulated data for intersections of road paths in the map data. In some embodiments, as described below, the mapping service uses the map data to automatically calculate intersection data for road access intersections. Such intersection data 4925 can be stored by indicating the type of intersection (e.g., point type, circular type) and a series of branches entering and exiting the intersection at different angles. Although map data 4915 and intersection data 4925 are shown as being stored separately, both data can be stored on the same physical storage device or on separate physical storage devices, and intersection data 4925 can actually be part of map data 4915. In addition, one or both of the map data and intersection data can be distributed across several physical storage devices (e.g., a series of disks for storing map data).
[0314] The map generator 4935 of some embodiments generates map information (e.g., map tiles) for transmission to a requester device. The requester device requests a map of a particular area (e.g., using latitude / longitude information), and the map generator 4935 creates (or uses pre-generated) map tiles for that area and then sends the data for those tiles (e.g., as encoded vectors and / or image data) to the device.
[0315] Route generator 4945 calculates an optimal route between two or more points in response to a user request. In some embodiments, route generator 4945 uses an optimization algorithm to calculate a route based on map data. A route can be defined as a series of intersections, a series of road paths, or in other ways. In addition, when a user requests a route, route generator 4945 provides intersection data for use by the device in turn-by-turn navigation. In some embodiments, intersection analyzer 4955 retrieves intersection data 4925 and modifies it for navigation of the route, as described below.
[0316] As shown, at stage 4910, device 4905 sends a request for a route to mapping service 4900. In some embodiments, the user enters a starting address (or location) and an ending address (or location), possibly including additional midpoint locations (e.g., starting at A, going to B, and then from B to C). The device then transmits the location information to the mapping service. In some embodiments, the device translates the location into latitude and longitude data, while in other embodiments, the mapping service performs this conversion.
[0317] At stage 4920, route generator 4945 accesses map data 4915 to generate one or more routes for the series of locations. In some embodiments, route generator 4945 uses an optimization algorithm to find the best (and second best, third best, etc.) route connecting the series of locations.
[0318] At stage 4930, the intersection analyzer 4955 identifies maneuvers along the route for which navigation directions need to be generated and retrieves intersection information for these maneuvers. Some embodiments generate turn-by-turn navigation directions to be provided to the device along with the route. To generate these directions, the mapping service 4900 identifies at which point the user following the directions must perform a maneuver (e.g., a right turn, a slight left turn, a U-turn, a merge, etc.) whenever the route changes path. In some embodiments, each of these path changes corresponds to a pre-tabulated intersection stored in the intersection data 4925. The intersection analyzer 4955 retrieves this intersection data. In some embodiments, each intersection is stored as a series of branch roads exiting the intersection at various angles (e.g., based on relative to north = 0°). In some embodiments, in addition to the intersection data, the route generator creates route planning directions that generally describe the maneuvers to be performed. Examples of such descriptions include "left turn," "exit highway ramp," "U-turn," etc. In other embodiments, this description is derived by the client mapping application based on the received intersection data.
[0319] Next, at stage 4940, intersection analyzer 4955 generates intersection information for the route design. In some embodiments, this results in modifying the angle to set the direction of travel into the junction to 0° (i.e., setting the branch of the route entering the junction to 180°). This effectively rotates the intersection description by the difference between true north and the incoming direction of travel of the route. In addition, intersection analyzer 4955 labels one of the branches as an exit branch. Some embodiments also label the entry branch, while other embodiments rely on the device to identify the 180° branch as an entry branch.
[0320] Stage 4950 shows that mapping service 4900 then transmits (e.g., via the same network that the device uses to transmit the route request) route data (i.e., route data and intersection data for navigation) to device 4905. Device 4905 then uses the intersection and route data generated by the mapping service to display navigation instructions for the user of the device, as shown at stage 4960. In some embodiments, the navigation instructions include displaying the intersection along with a stylized arrow that illustrates a maneuver through the intersection (in this case, a right turn).
[0321] Although mapping service 4900 is shown as including a map generator module and a route generator module, one of ordinary skill in the art will recognize that the mapping service may include additional modules or different subdivisions of the modules. The mapping service may consist of a single computing device (e.g., a server) that stores all functionality and data, or the functionality may be distributed among multiple servers (e.g., one process on a first server and a second process on a second server, multiple servers performing the same operation in parallel for different users, or other configurations of computing devices that perform the functionality described herein).
[0322] Figure 50 Conceptually, process 5000 is illustrated as being performed by a mapping service of some embodiments to generate and transmit route and intersection data to a user. As shown, process 5000 begins by receiving (at 5005) a request for a route between two locations on a map. In some embodiments, when a user requests a sequence of more than two locations, each segment is treated as a separate route (i.e., a first route from point A to point B, and a second route from point B to point C).
[0323] The process then uses the map data to generate (at 5010) at least one route between the locations. In some embodiments, the process uses an optimization algorithm to identify the best (or second best, third best, etc.) route between the two locations. These routes can be described as a series of vertices along a path, a series of intersections between paths, or using another description.
[0324] With the routes generated for the start and end locations, the process 5000 selects (at 5015) one of the generated routes to create turn-by-turn instructions for the route. The process then identifies (at 5020) the maneuvers to be performed along the route. In some embodiments, each time the route changes course, the mapping service identifies at which point the user following the directions will have to perform the maneuver (e.g., turn right, turn slightly left, U-turn, merge, etc.).
[0325] Next, the process retrieves (at 5025) an intersection for each maneuver. In some embodiments, each of these pathway changes corresponds to a pre-tabulated intersection stored by a mapping service. The generation of these intersections is described in detail below. In some embodiments, each intersection is stored as a series of branches exiting the intersection at various angles (e.g., based on relative to north = 0 °). In addition, in some embodiments, the intersection data stores the type of intersection (e.g., point type, circular type, traffic plaza, etc.).
[0326] The process then modifies (at 5030) the intersection information for each maneuver. In some embodiments, this results in modifying the angle to set the direction of travel into the junction to 0° (i.e., setting the branch of the route entering the junction to 180°). This effectively rotates the intersection description by the difference between true north and the incoming direction of travel of the route. In addition, some embodiments label one of the branches as an exit branch. Some embodiments also label the entry branch, while other embodiments rely on the device to identify the 180° branch as an entry branch.
[0327] The process 5000 then determines (at 5035) whether there are additional routes for which maneuver / intersection information is generated. If there are additional routes, the process returns to 5015 to select the next route. Different variations of routes from the start location to the end location may partially overlap, in which case some embodiments reuse the already generated intersection set for the overlapping portions.
[0328] Once intersections are generated for all routes, the process transmits (at 5040) the route and intersection information to the requester (e.g., the requesting device). As mentioned, in some embodiments, the requesting device uses this information to generate turn-by-turn navigation, including stylized junction / maneuvering arrows.
[0329] 2. Pre-calculation of intersection data
[0330] As described above, some embodiments pre-compute intersection data from stored map data (e.g., road segment and junction data). The mapping service then stores this intersection data for use in generating turn-by-turn navigation instructions. The following sections describe several processes for generating such intersection data, in which the mapping service receives vendor-provided junctions, identifies whether any set of received junctions should be merged into a single intersection, identifies pairs of road segments that should be joined together within the intersection, and generates angles for the intersection. Within this section, the term junction will generally be used to refer to vendor-provided information about where two path segments intersect, while the term intersection will refer to data generated by the junction that represents where two or more roads meet in the real world. Thus, in some embodiments, multiple junctions may be merged into a single intersection, and a junction between two road segments that is actually just a continuation of a single road may not be considered an intersection at all.
[0331] The following represents pseudo code for generating intersection data for a point-type intersection according to some embodiments:
[0332] Identify all inner road segments;
[0333] Identify all inner turn tunnels and mark them as inner sections;
[0334] For each inner road segment:
[0335] Collect all consecutive inner road segments;
[0336] Mark the collected inner road segments as processed;
[0337] The intersection is constructed by this collection of inner road segments;
[0338] In addition to other data (e.g., the locations of parks, waterways, businesses, etc.), map data also stores information related to pathways (i.e., roads, walkways, bike paths, etc.). In some embodiments, each pathway is received from a map provider as a series of segments (e.g., road segments). For a given road segment, in some embodiments, the stored data includes the segment's start and end junctions, geometry data defining the route taken by the path between the start and end junctions, a path representation (or "road form"), direction of travel (which, in some embodiments, may involve a one-way sign), one or more names for the path (e.g., "I-405S," "San Diego Freeway," etc.), a category indicating the path's level of importance, and a network class (the path connectivity graph to which the path belongs). In some embodiments, the geometry information includes a series of latitude / longitude vertices that the path traverses. In some embodiments, road form attributes include the following representations: one-way lane, two-way lane, motorway, fork, connecting road, walkway, stairs. Some embodiments may include additional representations.
[0339] Figure 51 Conceptually illustrated is a process 5100 of some embodiments for determining path segments between a set of junctions that should be treated together as a single intersection. As shown, process 5100 begins by receiving (at 5105) junctions between at least two path segments (e.g., road segments). In some embodiments, a mapping service receives (e.g., as pre-computed data from a map provider) a set of road segments and a set of junctions. Each road segment follows a path between two such junctions, and each junction references at least two road segments that enter and / or exit the junction. On the other hand, in some embodiments, junctions are not received from a map data provider, and the mapping service traverses the path data to identify intersections between paths and analyzes these intersections to pre-tabulate the junctions.
[0340] The process then determines (at 5110) whether any of the received path segments at the junction are dual carriageways. As mentioned, dual carriageways are a path representation used in some forms of map data. Many roads that are divided (e.g., by a median, double yellow lines, etc.) are received (and drawn) as two separate path segments, one for each direction. Each of the path segments is then labeled with the direction of travel and as half of a dual carriageway. Because users typically perceive the intersection of two roads, both divided by a median, as a single intersection (rather than four separate intersections), the junction generation process attempts to unify these four received junctions into a single intersection for presentation to the user for navigation purposes.
[0341] When no path segments are marked as dual carriageway, the process calculates (at 5115) the intersection legs using only the path segments specified in the received join (i.e., the intersection will only include one received join). In some embodiments, the calculation of the join legs results in the calculation of the angle from the join location for each road segment specified for the join. The process then ends. Figure 52 An example of such a junction 5200 is shown, also illustrating that it is not required that path segments meet at right angles or that the path continue in a straight line through the junction.
[0342] When at least one of the path segments specified for the received join point is a dual carriageway, the process determines (at 5120) whether there is a cross-traffic turn exiting the dual carriageway at the join point. A cross-traffic turn is a turn exiting the dual carriageway in the direction of the matching half of the dual carriageway (i.e., the other direction of the road), if a cross-traffic turn exists. In the United States, a left turn is a cross-traffic turn. While the examples shown in this document relate to right-hand drive (i.e., driving on the right side of the road), those of ordinary skill will recognize that the examples are equally applicable to left-hand drive areas (e.g., England). Figure 53 An intersection 5300 is shown that includes two dual-lane paths 5305 and 5306 and a one-way road 5310. At junction 5315, there is no turning exit for the dual-lane crossing traffic, as the only options are a right turn exiting the dual-lane path 5305 and a left turn exiting the one-way street 5310. In the absence of such a turn, process 5100 stores (at 5125) the received junction, while recognizing that the received junction may still be part of a larger intersection, in order to determine whether to include the received junction in the larger intersection along with other received junctions (e.g., junction 5320 between the one-way road 5310 and the dual-lane path 5306). For example, at intersection 5300, the process may wish to join received junction 5320 with received junction 5315 into a single, larger intersection. The process then ends.
[0343] When there is a two-lane crossing traffic turn away at a junction (e.g., at junction 5320), the process moves in the direction of the crossing traffic turn (at 5130) until the next two-lane path is reached. In some embodiments, because road segments start and stop at junctions, the next two-lane path will be reached at a different received junction (but not necessarily the next junction if a road, such as a left-turn lane, is received as a separate path segment). For example, from intersection 5320, the process will traverse path 5310 away from junction 5320 until the next two-lane path is reached at junction 5315.
[0344] The process 5100 then determines (at 5135) whether the two-lane path arrived at 5130 has a direction of travel that is in the opposite direction of the source two-lane path. This is essentially a quick determinant of whether the second two-lane path is likely a matching path for the first two-lane path (i.e., whether they are likely two sides of the same road). In most cases, due to the nature of how roads are typically constructed, this next two-lane path will be a matching path.
[0345] In the event that the second dual carriageway is not in the opposite direction of the source path, the process proceeds to 5125 to store the newly arrived junction for later use in determining whether to include the junction with any other received junctions. For example, if the left turn of path 5306 leaves onto another dual carriageway with a downward travel direction, path 5306 can be assumed to have no data match (at least with respect to the junction), but the newly identified path itself may have a match.
[0346] On the other hand, if the two dual carriageways have opposing directions of travel, the process identifies and stores (at 5140) the road segment traversed by the crossing traffic turn. Figure 53 In the example of , the road segment from junction 5320 to junction 5315 will be stored. In some embodiments, this road segment will be used as part of the additional junction merging process. The process then ends.
[0347] The above process 5100, when applied to all junctions within the map area, will generate a set of road segments between the dual carriageways. Some embodiments use these road segments to link received junctions together and identify additional received junctions to include in a single intersection definition. The following represents pseudo code for some embodiments for identifying all such "inner" road segments of a complex intersection:
[0348] For each road section with dual carriageways;
[0349] For each link with a crossing traffic turn, where a path can be assembled to the other side of the intersection;
[0350] Mark all sections of the path on the other side as inner sections;
[0351] This pseudo code includes a determination as to whether a path can be assembled from a road segment to the other side of an intersection. The following includes pseudo code for such a determination in some embodiments:
[0352] Add the first segment to the path;
[0353] Get the connection from the last segment on the path;
[0354] Iterate over each connection to find the connection to the other side or the connection that is the best continuation;
[0355] If the connection is the other side, the marking is successful and ends;
[0356] If there is no connection to the other side, and no connection to the best continuation, then
[0357] The annotation fails and ends;
[0358] otherwise:
[0359] Add a segment to the end of the path;
[0360] If the path is too far, the labeling fails and ends;
[0361] If there are too many intersections, the labeling fails and ends;
[0362] Otherwise, return to obtain the connections of the added road segments and iterate over the connections;
[0363] Figure 54 A process 5400 is conceptually shown for joining several junctions together into a single intersection and identifying branches of the intersection. The process 5400 begins by receiving (at 5405) a set of intersecting road segments between two lanes. In some embodiments, a process such as Figure 51 The mapping service then groups together sets of such road segments that intersect each other (ie, at the received join points). Figure 55 A co-existing intersection 5500 is shown between a dual carriageway having paths 5505 and 5506 and a dual carriageway having paths 5510 and 5511. In this figure, the set of intersecting road segments is shown as a thicker line.
[0364] The process then identifies (at 5410) all junctions and path segments that are directly connected to the set of intersecting road segments at the junction. That is, the set of intersecting paths intersects at the junction, but these junctions may contain additional path segments. For example, in exemplary intersection 5500, the eight dual-lane path segments exiting the intersection all intersect with inner (thicker) path segments at four junctions. Therefore, the four junctions and eight outer path segments are included in the intersection.
[0365] Figure 56 On the other hand, an intersection 5600 is shown in which the left-turn lanes are defined as separate path segments. In this case, the left-turn lanes are identified by process 5400 because they intersect the inner road segment at the junction point received in the initial map data. In most cases, the left-turn lanes can be characterized as slip roads or single-lane roads in the map data.
[0366] The following represents pseudocode for some embodiments for identifying all turning lanes that are considered to be "inside" an intersection:
[0367] For each road section with dual carriageways;
[0368] For each link with a crossing traffic turn, where a path can be assembled to an inner segment;
[0369] Mark all segments on the path to the inner segment as inner segments;
[0370] The pseudo code includes a determination as to whether a path can be assembled from a road segment (e.g., a turning corridor) to an inner road segment. The following includes pseudo code for such a determination in some embodiments:
[0371] Add the first segment to the path;
[0372] Get the connection from the last segment on the path (i.e., connect to the last segment at the junction).
[0373] road section);
[0374] Iterate over each link to find the inner link or
[0375] Find the connection that is the best continuation;
[0376] If the connection is an internal road segment, the marking is successful and ends;
[0377] If no link is inside and no link is best
[0378] If it continues, the mark fails and ends;
[0379] otherwise:
[0380] Add a segment to the end of the path;
[0381] If the path is too far, the labeling fails and ends;
[0382] If there are too many intersections, the labeling fails and ends;
[0383] Otherwise, return to get the connection of the added segment and perform
[0384] Iterative processing;
[0385] Next, the process 5400 defines (at 5415) a new intersection as including all identified junctions and path segments, including those that directly intersect the initial set of path segments. Figure 56 In the case shown, this would include a junction where the left-turn lane leaves its source two-lane road segment and a left-turn lane that intersects the initial road segment. In this case, some embodiments identify a fork or other junction of the one-lane path segment (i.e., a starting junction), which would be where the path segment intersects one of the two-lane path segments before entering the intersection. When a one-lane path segment remains within a threshold distance (e.g., 1 km) of a (hypothetical) two-lane path pair, some embodiments assume that the path segment is part of the road defined by the two-lane path and eliminate consideration of the junction.
[0386] When dealing with slip roads or other connecting roads outside the dual carriageway (e.g. Figure 57 5705 shown in intersection 5700 of FIGURE 5700 ), some embodiments do not treat the turnoff as a path into the two-lane intersection. Instead, some embodiments identify the path characterization as a turnoff and attempt to form a closed loop that includes the start and end junctions of the turnoff. When this closed loop shares a common junction with the newly defined intersection (as is typically the case), the turnoff may be associated with the intersection but is not treated as an internal path of the intersection. On the other hand, when the newly defined two-lane intersection has grown due to, for example, the presence of a left-turn lane such that the turnoff junction is surrounded by the intersection that now includes intersecting one-way lanes (as for FIGURE 5701 ). Figure 58 In the case of a turnout 5805 in the newly defined intersection 5800, some embodiments treat the turnout as being inside the newly defined intersection. In the description of the intersection, these left turns, turnouts, etc. will typically be omitted because the user typically does not want complex instructions, but instead will want an instruction to "turn right onto San Vicente Blvd" or some similar instruction.
[0387] With the set of road segments and junctions that form the defined intersection, the process entails merging the two-way road into a single junction branch. Process 5400 next defines (at 5420) the set of all paths that enter the intersection, and defines (at 5425) the set of all paths that exit the intersection. For a two-way road that is a one-way path (half of a two-way road), the path will typically have an exit edge and an entry edge. For merging purposes, some embodiments treat each road segment (the one exiting the intersection and the one entering the intersection) as a separate path. A one-way road that is not inside a two-way road (e.g., Figure 59 The additional bidirectional path 5905 in the intersection 5900 will typically be treated as a separate branch and in some embodiments is not part of the merge analysis.
[0388] Next, the process determines (at 5430) whether the entry path set includes any unpaired two-way roads. If no such paths remain in the set (or no such paths exist at the first location), the process stores (at 5435) any unpaired two-way roads remaining in the exit path set as separate branches of the junction. This typically occurs with mislabeled map data (a road that is actually a one-way street) or with overly strict merging criteria (leaving unmerged entry and exit path pairs).
[0389] When there are unpaired entry paths, the process selects (at 5440) one of the entry paths. The process then determines (at 5445) whether there is a potential match in the exit set. In some embodiments, a potential match is a dual carriageway found by traversing the previously identified road segment to the left (or right in the case of a left-hand driving zone) or traversing the intersection in a clockwise manner.
[0390] When there is no potential match (e.g., the next identified dual carriageway in the traversal is also an entry path, or the exit set is empty), the process stores (at 5450) the entry path as a separate branch of the intersection and then returns to 5430 to find the next unpaired entry path. On the other hand, when there is a potential match, some embodiments determine (at 5455) whether the potential pair meets a set of dual carriageway matching criteria. In some embodiments, these are criteria used to determine whether a dual carriageway pair is actually two sides of the same road. Some embodiments determine (1) whether the two paths are within a threshold distance (e.g., 25m, 50m, etc.) of the paths entering / exiting the intersection, and (2) whether the angles at which the paths arrive at their junctions in the intersection are within a threshold range (e.g., 5°, 10°, etc.) of each other. To calculate the angle, some embodiments use the vertex closest to the edge of the intersection (or the location of the junction where a path segment intersects with other road segments in the intersection) and the vertex that is outside a certain predetermined distance (e.g., 50m). The process then calculates the angle relative to north for the line between the two vertices.
[0391] In some embodiments, the mapping service also considers the names of the paths to determine whether these match. When the names match, such embodiments may relax the geometry criteria for the matching pairs (i.e., allow a larger distance between the paths or a larger difference in the angles between the paths). Matching names might be, for example, "CA-1 South" and "CA-1 North," or if both paths include "Welsh Avenue" as one of their names. Some embodiments may also consider road class data to confirm that a dual carriageway is matched.
[0392] If the two paths match, the process merges the paths (at 5460) into a single branch of the newly defined intersection. As noted above, the intersection is stored as a collection of branches at different angles. For the merged paths, some embodiments store the angle as the average of the angles of the two paths that make up the branch. Figure 60 6000 is shown as reducing an eight-path intersection into four branches, where the angle of the right branch 6010 is half the offset of the right exit path 6005 from the horizontal because the right entry path is on the horizontal line. As shown conceptually, in some embodiments, directions (entry / exit) are not stored for the intersection branches. The mapping service uses the map data to generate a route that includes the intersection and the direction of travel of the road, so the route will not go down the wrong road on the branches of the intersection.
[0393] On the other hand, when the paths do not match, the process stores (at 5465) each of the paths as a separate branch of the intersection. Figure 61 A different eight-path intersection 6100 is shown reduced to five branches. In this case,
[0394] The two-lane paths 6105 and 6106 on the right do not merge and are therefore treated as separate branches 6110 and 6111 of the reduced intersection. In this example, the angle at which each of these branches exits the intersection is the angle stored for the branch (no averaging). The process 5400 then returns to 5430 to determine if any entry paths remain. As described, once the entry path set is empty, the process proceeds to 5435 and then ends.
[0395] The following represents pseudo code of some embodiments for generating data for an intersection once interior road segments have been identified for the intersection (e.g., operations performed by some or all of process 5400):
[0396] Collect all outer road segments that touch the inner road segment of the intersection; identify the outer road segments surrounded by the inner road segments in the intersection and mark them as inner;
[0397] Grouping together pairs of incoming and outgoing segments representing the same road;
[0398] Calculate the exit angle for each pair and for each unpaired road; construct a template intersection pattern with one branch for each angle; if a pattern exists for a previously generated intersection, use the existing pattern to save space (make the intersection refer to the existing pattern);
[0399] Otherwise if the pattern does not exist, a new entry for the pattern is created and stored;
[0400] As noted, some embodiments store each intersection as a data structure that indicates the branches of the intersection and the angles at which the branches enter and / or exit the junction. Figure 62Conceptually, an example of this type of data structure 6200 for a point-type intersection in some embodiments is shown. As shown, an intersection includes an intersection ID (which is a unique identifier in some embodiments), a map data association, and a branch road set containing angles and types. In some embodiments, the map data association associates the intersection data structure with the actual location in the map. In some embodiments, this is simply a latitude / longitude point, but in other embodiments, it can also be composed of other data (e.g., a list of path segments or aggregated paths that meet at the intersection). Each branch road includes a type and an angle. In some embodiments, the type is an intersection type. Some embodiments define two intersection types: point type and ring type. Other embodiments may then include additional intersection types, such as traffic plazas. Some embodiments store the intersection type as a characteristic of the intersection, rather than storing it separately for each branch road, but other embodiments recognize that the intersection is partially ring-type but partially a point-type intersection. Data structure 6200 includes four branches in the cardinal directions of 0° (north), 90° (east), 180° (south), and -90° (west). In some embodiments, the intersection data structure also includes references to any junctions (ie, data received from a map data provider) and path segments contained within the defined intersection. For a typical intersection of two dual carriageways, such a data structure involves four junctions.
[0401] Figure 63 The data structure 6300 for a roundabout of some embodiments is shown. Some embodiments provide specialized processing for roundabouts. The following represents pseudo code for generating intersection data for a roundabout of some embodiments:
[0402] Identify all circular road sections;
[0403] For each ring road segment:
[0404] Collect all consecutive circular segments;
[0405] Mark the collected circular road segments as processed;
[0406] Construct a ring through this collection of ring segments
[0407] intersection;
[0408] In some cases, the map data identifies roundabouts (e.g., as a form of road or by another indicator). This allows the mapping service intersection calculator to begin its specialized automatic roundabout processing. Specifically, when performing roundabout processing, the mapping service attempts to identify extended connecting road pairs (i.e., portions of the road that extend into and out of the roundabout). In some embodiments, the intersection calculator traverses the roundabout (e.g., in a counterclockwise manner for right-hand driving), looking for exit paths that are followed by entry paths within a specific distance (e.g., an angular distance). The process then determines whether to combine these paths, taking into account factors similar to those for merging into a dual carriageway at a point-type intersection. For example, the factors used might include whether the names are similar, whether the distance between the exit / entry paths is small enough, and potentially other factors. As a result of this processing, when a random road intersects a roundabout between structures that would otherwise appear to be an entry / exit combination, some embodiments treat this as three separate branches.
[0409] To calculate the angle of the leg, some embodiments determine where the leg intersects the roundabout, rather than the approach angle of the road.For entry / exit combinations, the process takes the average of the two paths. Figure 64 The reduction of a roundabout 6400 to intersection data is conceptually illustrated. The top path, despite approaching at an angle of approximately 30° relative to North, is designated as a 0° leg—the user is primarily interested in the distance around the traffic circle at the intersection, rather than the angles at which they enter and exit. The other three legs are also assigned cardinal directions, as their extension toward these directions ultimately balances out. Data structure 6300 illustrates the data structure for a roundabout junction 6400. However, other embodiments use the angle at which a path enters or exits a roundabout, rather than the distance around the roundabout at which the path intersects.
[0410] The following represents pseudo code of some embodiments for generating data for a roundabout once the roundabout sections have been identified for the intersection:
[0411] For a collection of circular road segments forming a single loop:
[0412] Collect all non-circular road segments that touch the loop, sorted by the direction of travel around the loop;
[0413] Grouping together pairs of consecutive circular exit / entry segments representing the same road;
[0414] Assign angles to each paired and unpaired road segment;
[0415] Subtract the smallest angle from all angles (so the smallest angle = 0); construct a template intersection pattern with one branch for each angle; if a pattern exists for a previously generated intersection, use the existing pattern to save space (make the intersection refer to the existing pattern);
[0416] Otherwise if the pattern does not exist, a new entry for the pattern is created and stored;
[0417] As pointed out by the above-mentioned example of pseudocode, some embodiments perform additional compression when storing intersections.The real world comprises millions (or hundreds of millions) of independent intersections, but many of these intersections have identical configurations (especially when allowing very small angle variations).Therefore, some embodiments use compression in the process of storing intersections, rather than storing independent data for each of hundreds of millions of intersections.When processing each intersection, some embodiments store template pattern for this intersection.When identifying additional intersections with template pattern, this type of embodiment stores reference to this pattern (and still creates independent data structure simultaneously, because for two intersections following identical pattern, position information is different).
[0418] 3. Modification of navigation junction data
[0419] The previous section described the generation of complex intersection data, which is typically performed as an offline process before route generation. However, during route generation, some embodiments modify the intersection data for transmission to the user. The mapping service providing the route data modifies the angles to make them relevant to the incoming direction and marks one of the branches as an outgoing branch.
[0420] Figure 65 Conceptually, a process 6500 for modifying intersection data in order to provide navigation information for a route is shown in some embodiments. As shown, the process begins by receiving (at 6505) a route for which intersection information is generated. As described above, some embodiments generate one or more routes for each set of start and end locations requested by a user device. Each of these routes consists of a series of maneuvers at various path intersections (i.e., at road intersections).
[0421] As shown, after identifying a route, process 6500 selects (at 6510) the next intersection along the route. Some embodiments begin with the first intersection starting from the starting point of the route (i.e., the first maneuver that the user will perform along the route). Many routes involve long stretches along a particular road, traveling straight through multiple intersections (possibly including junctions of two sections that are part of the same road and do not intersect with other roads). In some embodiments, navigation instructions do not include information about intersections where no turning maneuvers will be performed. Therefore, the next intersection is actually the next intersection along the route where the maneuver will be performed.
[0422] The process then retrieves (at 6515) the pre-computed intersection data as a set of branches with associated angles. As described above, some embodiments store a data structure for each intersection that lists the branches of the intersection and the angle of each branch. Figure 62 and Figure 63 Examples of such data structures are shown for both point-type intersections and roundabout-type intersections.
[0423] After retrieving the data structure for the selected intersection, the mapping service rotates the intersection definition to normalize the definition to the direction of the route entering the intersection. Thus, process 6500 identifies (at 6520) the entrance leg of the intersection and sets the entrance leg to a predetermined angle. Some embodiments set the direction of movement into the intersection to 0° and therefore set the entrance leg of the intersection to 180°.
[0424] The process then rotates the other branches of the intersection. As shown, the process selects (at 6525) the next branch of the intersection. In some embodiments, the branch and angle are stored in an array, list or similar data structure, and the process traverses the data structure. The process sets the angle of the selected branch (at 6530) based on the angular distance from the entrance branch. For example, if the entrance branch is stored as 0 ° (i.e., pointing to the north), the branch stored as 95 ° will offset 180 ° to -85 °. In addition, the process determines (at 6535) whether the selected branch is the exit branch of the junction (i.e., the branch where the route exits the intersection). In order to correctly display the turn-by-turn navigation instructions at the client mapping / navigation application, the device needs to know which branch of the intersection the route exits. Therefore, when the selected branch is the exit branch, the process 6500 marks (at 6540) the selected branch as such. The process then determines (at 6545) whether any additional branches of the intersection are still to be transformed for the route. If there are still additional branches, the process returns to 6525 to select the next branch of the junction. When all branches have been processed for the current intersection, the process 6500 determines (at 6560) whether there are still additional intersections along the route that need to be modified. If there are still additional intersections, the process returns to 6510 to select the next intersection. When the last intersection has been modified, the process ends.
[0425] Figure 66 Shown is a conceptual drawing of a route through an intersection 6600, a data structure 6605 for the intersection, and a modification of the data structure to create a new data structure 6610 for turn-by-turn navigation instructions. As shown, the route enters from the right (90 ° branch) and exits the intersection at the bottom (-162 ° branch). In the modified data structure, the entrance branch has been rotated to 180 °, resulting in a 90 ° rotation of the other branches. The branch at 18 ° rotates to 108 °, the branch at -65 ° rotates to 25 °, and the branch at -162 ° rotates to -72 °. In addition to the rotation angle, data structure 6610 also has a last branch that is marked as an exit for navigation. Some embodiments include a binary exit field, with the exit branch marked as "1" and all other branches marked as "0."
[0426] D. Generation of land cover geometry
[0427] In addition to generating road geometry for map tiles, some embodiments also generate land cover geometry. Much like road segment data is received from various sources, data describing land cover can also be received (e.g., as a series of vertices indicating the intersection of a particular land cover body). Land cover can include bodies of water (e.g., rivers, oceans, lakes, swimming pools, etc.), political entities (e.g., the intersection of states, countries, cities, parks, etc.), regional designations (e.g., rural / urban / suburban, desert / mountains / forest, etc.), or other data describing the land between roads. First, some embodiments use these coordinates to grow geometry for the land cover project.
[0428] After growing the geometry for the land cover, some embodiments use various operations to resolve intersections between the geometries. For example, when combining data from different sources, the location data indicating object intersections may not be perfectly aligned and therefore there may be gaps between object geometries or overlaps of geometries. Some embodiments use different operations to resolve intersections between different geometries, depending on the type of object. Additionally, the land cover geometry processing uses road segment data to fill in the land cover geometry and ensure that no gaps are left between the land cover and the road. Some embodiments grow the land cover geometry toward the road beyond its marked intersection, stopping the geometry at the road centerline location. While this creates an overlap between the land cover geometry and the road geometry, in some embodiments the client mapping application includes instructions to render the road geometry on top of the land cover geometry. The following subsections describe some of these different methods for resolving differences or mismatches between various land cover intersections.
[0429] 1. Incorporate land cover into roads
[0430] Figure 67 A process 6700 is conceptually illustrated that some embodiments of the present invention perform to merge land cover polygons into road polygons. In some embodiments, the process of merging land cover polygons is performed to resolve data mismatches that may arise from receiving land cover polygon data and road polygon data from different data sources. Figure 68 to describe the process of merging land covers, Figure 68 An example of merging land cover into road polygons in three stages 6801-6803 is shown.
[0431] like Figure 67As shown, process 6700 receives (at 6705) a road polygon. Process 6700 then receives (at 6710) at least two land cover polygons. In some embodiments, the road polygons and the land cover polygons are received from two different data sources, while in other embodiments, they are received from the same data source. Figure 68 Stage 6801 in FIG. 7 shows an example of two such land cover polygons 6810 and 6820. In this example, a different land cover type is represented by each land cover polygon. For example, land cover 6820 may be a lake, while land cover 6810 may be a park. However, in other embodiments, land cover polygons 6810 and 6820 may be of the same type. Furthermore, the process may receive more than two land cover polygons, which is not shown for simplicity. Stage 6801 also shows the empty space between the two land cover polygons. Since land cover represents all areas except roads, the empty space is assumed to represent where roads typically extend. However, due to the irregularity of the space between the two land cover polygons 6810 and 6820, if a road were drawn between the two land cover polygons, there would be an irregular empty space between the road and the land cover that would be visually unpleasing when the map is rendered on the user's device. Therefore, process 6700 merges the land cover polygons into the road, as described below.
[0432] The process 6700 draws (at 6715) a road polygon in the empty space between two land cover polygons, as shown in FIG. Figure 68 68. In some embodiments, the road polygon represents a road located between two land cover polygons. In some embodiments, the road may not fit neatly in the center of the empty space. Portions of the road, or the entire road, may overlap one or both polygons. For simplicity, process 6700 is described assuming that the road lies in the space between the land cover polygons.
[0433] Next, the process 6700 grows (at 6720) the land cover polygon up to the road polygon. Figure 68 As shown in stage 6802, the process expands the left side of land cover polygon 6810 until the polygon reaches the right side of road polygon 6830. The process also expands the right side of polygon 6820 until it reaches the left side of road polygon 6830. In some embodiments, when the road polygon is not entirely located in the center space between land cover polygons, the process alternatively grows the land cover polygon to the center of the gap. In other embodiments, the land cover polygon grows beyond the road geometry, covering part or all of the road geometry.
[0434] To perform this expansion of step 6720, in some embodiments, the process rasterizes the polygons before expansion. By rasterizing the polygons, the process generates pixel data for each pixel within each land cover polygon. The merging process uses this pixel data to fill adjacent pixels with similar data. For example, if a null pixel is adjacent to a pixel of a land cover polygon that is a type representing a paved area, the null pixel is also changed to represent the same paved area type. The process continues to grow the null pixels in this manner until they reach a road polygon 6830, which acts as an obstacle and terminates the polygon expansion process.
[0435] although Figure 68 An example is shown in which only one road polygon is drawn, but other embodiments include multiple road polygons representing more than one road between two land cover polygons. In such cases, empty space may remain between the two roads because the land cover polygon does not expand into the space between the two roads (i.e., the road acts as an obstacle to polygon expansion).
[0436] When empty space remains between the two roads, process 6700 marks the empty space between the roads as filler (at 6725). In some embodiments, the marked filler is interpreted and rendered according to different instructions on the client application. For example, in some embodiments, the user device renders the empty space as a road divider. In other embodiments, the user device renders the empty space as green space, asphalt, etc.
[0437] After marking any empty space that may exist between the two roads, some embodiments convert the newly added pixels into vectors to reduce the number of vertices from the land cover polygon, and the union of the added pixels with the original receiving polygon from step 6710 of process 6700. The process then draws (at 6730) the road on the road polygon. In some embodiments, the drawn road may also overlap polygons that are close to the road polygon. Figure 68 The third stage 6803 shows operation 6730 of process 6700, in which the process draws a road 6840 on the road polygon between two polygons 6820 and 6810. In this example, road 6840 covers the road polygon and overlaps portions of polygons 6810 and 6820.
[0438] Next, the process optionally performs (at 6735) polygon optimization on certain land cover polygon types to reduce the number of vertices in the land cover polygons. In some embodiments, this process involves removing certain land cover polygons of a particular type and drawing the land cover polygons on a separate layer. For example, this optimization technique can be performed on a small body of water such as a swimming pool (or pond), which generates a large number of vertices within the polygon corresponding to the pool and its surrounding area. For example, four swimming pools surrounded by paved areas are represented as four polygons to represent the pools themselves. In addition, the paved areas are represented as polygons with four holes within them. In some embodiments, the polygons corresponding to the swimming pools and their corresponding holes are removed and placed in a separate layer. The paved areas are drawn and the pools are drawn on top of them. In this way, the number of points associated with the pools is reduced.
[0439] 2. Merge boundaries based on the reliability or quality of source data
[0440] Figure 69 Process 6900 is conceptually illustrated as merging intersections between adjacent polygons received from two different sources in some embodiments. In some embodiments, process 6900 is performed to resolve mismatched intersection coordinates received from different sources, where one data source is deemed more reliable than the other. Figure 70 and Figure 71 Describes the process of merging boundary coordinates. Figure 70 Two stages 7001 and 7002 of some embodiments of the present invention are shown that resolve identified gaps 7010 , 7040 , and 7060 by merging boundary coordinates between adjacent polygons 7020 and 7030 . Figure 71 Two stages 7101 and 7102 of merging the boundary coordinates of polygons 7120 and 7130 surrounding identified overlaps 7110 and 7140 are shown in some embodiments of the present invention.
[0441] like Figure 69 As shown, process 6900 receives (at 6905) two polygons of two different land cover types adjacent to each other. For example, the two land cover types may represent an ocean and a park. The process may also receive several other land cover types representing other map features. Next, the process identifies (at 6910) any gaps and overlaps between the intersections of the two polygons.
[0442] like Figure 70 As shown in stage 7001 in FIG. 7 , there are gaps 7010, 7040, and 7060 between polygons 7020 and 7030. Figure 71As shown, there are overlapping regions 7110 and 7140 between polygons 7120 and 7130. Although shown as separate figures for simplicity, some embodiments of the present invention may merge polygon intersection coordinates around both gaps and overlaps that occur between the same two polygons during the merging process.
[0443] Re-reference Figure 69 , after identifying gaps and overlaps between polygon intersections, process 6900 determines (at 6915) which polygon has intersection coordinates from a more reliable source, and therefore has more reliable intersection coordinates. For example, if polygon 7120 represents an ocean and polygon 7130 represents a park, and the vendor or source providing the ocean polygon is deemed a more reliable or better quality source, the process resolves (at 6920) the overlap in a manner that favors the more reliable data. In some embodiments, the process resolves overlaps in a manner that favors data representing particular types of polygons. For example, in the ocean / park example described above, the process may determine that the ocean polygons are more reliable than the park polygons and resolve the overlap in a manner that favors the more reliable ocean polygon data. Other embodiments resolve overlaps in a manner that favors polygons of larger or smaller size. For example, some embodiments resolve overlaps in a manner that favors polygons with the largest perimeter or largest surface area, while other embodiments resolve overlaps in a manner that favors polygons with the shortest perimeter or smallest surface area. As in Figure 71 As shown in stage 7102 of , overlapping regions 7110 and 7140 are removed by matching the intersection of the less reliable polygon 7120 to align with the intersection of the more reliable polygon 7130. In other words, the coordinates of the more reliable polygon 7130 are used to determine the intersection between the two regions.
[0444] To resolve the gaps, for polygons with less reliable intersection coordinates (or lower quality data), the process calculates (at 6925) a threshold based on the shape of the polygon. In some embodiments, the shape of the polygon is based on the area of the polygon divided by the perimeter of the polygon multiplied by a scaling factor. When more than a predetermined number of intersection points on the polygon with the less reliable coordinates are closer to the more reliable polygon than the threshold, the process uses (at 6930) the intersection coordinates of the more reliable polygon and extends the less reliable area to join the more reliable area. Once the distance between two polygons is greater than the threshold, the two polygons remain separated. For example, if Figure 70If polygon 7030 in FIG1 is from a more reliable source (or has better data quality), gaps 7010 and 7060 are closed (which, in this example, is less than the calculated threshold) by extending the boundary of the less reliable region 7020 to the boundary of region 7030. In other words, the boundary coordinates of region 7030 are used for both regions in the region of gaps 7010 and 7060. On the other hand, in this example, the distance between the two regions in a portion of gap 7040 is greater than the calculated threshold. Therefore, gap 7040 is only partially closed.
[0445] In these examples, gaps and overlaps are merged in a manner that favors more reliable intersection data. However, other embodiments may merge intersection coordinates around gaps and overlaps in other ways, including merging in a manner that favors polygons or intersection data with the most detail.
[0446] 3. Merge regional boundaries based on boundary data analysis
[0447] Figure 72 A process 7200 is conceptually illustrated, which is performed by some embodiments of the present invention to resolve boundary conflicts between regions. A region can be any land cover area, such as a body of water, a park, a building, etc. A region can also be an administrative area, such as a state, a city, a county, a country, etc. The process extracts boundaries to resolve mismatches when two data sets are given a common boundary. Such mismatches can occur, for example, when boundary data for regions that share a common boundary is obtained from different sources. Process 7200 attempts to resolve mismatches by favoring a single boundary.
[0448] like Figure 72 As shown, process 7200 receives (at 7205) polygons corresponding to different regions. The process then identifies (at 7210) overlaps between the regions. The process resolves the overlap by including (at 7215) the overlapping region in the region with the boundary that retains the most boundary data points. In some embodiments, step 7215 involves subtracting the first polygon from the second polygon or vice versa. Some embodiments resolve the overlap based on which of the aforementioned operations has more boundary data points, thereby retaining more boundary detail.
[0449] Next, process 7200 determines (at 7220) whether all overlapping regions have been examined. If not, the process returns to step 7210 to identify another overlapping region between different regions.
[0450] After all overlapping regions have been checked, process 7200 continues to step 7225 to identify gaps between adjacent regions. Some embodiments of the present invention obtain a list of gaps between adjacent boundaries by taking the union of all regions and subtracting all regions that have been checked for overlap.
[0451] Once the process has identified a gap between adjacent regions, process 7200 determines (at 7230) whether the gap is between only two regions. If so, the process uses (at 7235) the boundary of the region with the greatest number of points along the intersection. Otherwise, the gap is between three or more regions. In this case, the process uses (at 7240) the boundary of the region with the longest boundary. However, in some embodiments, if the gap is between four or more regions, the process may flag the gap for human analysis. The process then determines (at 7245) whether all gaps have been examined. If not, the process returns to step 7225 to identify another gap. If so, the process ends.
[0452] Figure 73 An example of resolving a boundary conflict between two adjacent regions 7350 and 7360 by executing process 7200 is shown. Although for simplicity Figure 73 Only two regions are shown, but process 7200 can be performed for mismatched boundaries between several regions. As shown, regions 7350 and 7360 share a common boundary 7390. However, a boundary conflict exists between the two regions. Figure 73 The first stage 7301 shows a broken portion 7340 of the boundary 7390 between regions 7350 and 7360. The broken portion 7340 shows an example of a boundary conflict, such as an overlap 7310 and a gap 7330, which can be resolved by performing the process 7200.
[0453] Figure 73 The second stage 7302 shows a broken portion of what the boundary 7370 may look like after the process 7200 is performed on the boundary 7390. Figure 72 , the process 7200 will first identify (at 7210) the overlap 7310 and resolve the conflict by including (at 7215) the portion of the region that retains the most bounding data points. Figure 73 In the example shown, process 7200 determines that boundary 7395 around region 7360 holds more boundary data points and includes that boundary as shown in stage 7302. Process 7200 then identifies (at 7225) that gap 7330 must be resolved. The process also attempts to remove gap 7330 between regions 7350 and 7360 by using (at 7235) the boundary that holds more boundary data points. Figure 73 In the example shown, process 7200 determines that for that portion of the boundary, more boundary data points remain around boundary 7397 of region 7350 and include the boundary as shown in stage 7302. In some embodiments, some or all of the operations defined in processes 67, 69, and 72 are performed by one or more software modules.
[0454] E. Tile Generation
[0455] One of the last operations performed by the set of servers generating map tiles is a tile cut operation, which generates map tiles and encodes the geometry and vector data in those tiles. In some embodiments, the set of servers defines the map as a tile tree having multiple levels corresponding to the multiple zoom levels used to view the map. For example, in some embodiments, the tile tree data structure is a quadtree having twenty-one levels corresponding to the twenty-one zoom levels used to view the map. In the quadtree, each node is a map tile, and each non-leaf node has four child tile nodes.
[0456] The tile cutting operation has several novel features. One novelty is the fast mechanism it uses to associate each geometry generated for a map with one or more tiles in the tile quadtree. In some embodiments, this fast mechanism uses distributed computing to quickly associate a large number of geometries in a map with a large number of map tiles. The task of determining whether each geometry should be associated with a tile is treated as a separate task from other similar tasks, each of which can be assigned to different computing resources in a distributed computing environment.
[0457] For example, a computational resource can determine whether a geometry intersects a tile. If so, a separate intersection operation is performed for each of the tile's subtiles to determine whether the geometry intersects the subtile. A separate computational resource can perform each different intersection operation independently of other computational resources. This process can be thought of as "throwing" the geometry down the tile quadtree, identifying each tile node that the geometry intersects, and then sending the intersecting portion of the geometry to each of the intersecting nodes' children to determine whether the child node intersects the geometry. The geometry identified for each tile node is then used in the tile cutting process to generate tiles.
[0458] Another novelty involves a fast mechanism for re-encoding road vector data within map tiles to remove unnecessary data. Again, the tile-cutting operation uses distributed computing to distribute the task of re-encoding the vector data. Each assigned task involves initially recording the road data for each tile (which may overlap) on the same canvas in order to merge overlapping road geometry. The task then involves re-encoding this data to generate non-duplicate vector data.
[0459] Such tile cutting operations of some embodiments are further described in U.S. Provisional Patent Application No. 61 / 657,872, entitled “Scalable and Efficient Cutting of Map Tiles,” filed on June 10, 2012; U.S. Patent Application No. 13 / 632,003, entitled “Scalable Processing for Associating Geometries with Map Tiles,” filed on September 30, 2012; and U.S. Patent Application No. 13 / 632,004, entitled “Scalable and Efficient Cutting of Map Tiles,” filed on September 30, 2012. Provisional Application No. 61 / 657,872 and Non-Provisional Applications Nos. 13 / 632,003 and 13 / 632,004 are incorporated herein by reference.
[0460] III. Client-side Generation of 3D Presentations
[0461] The mapping application of some embodiments uses various novel techniques to present a 3D display of a map while the map is being browsed or while the map is providing a navigation display. For example, as described above, the mapping application renders the 3D display based on the vantage point of the virtual camera, and uses various methods of automatically moving the virtual camera (i.e., moving the perspective rendering position) in certain situations to achieve the desired 3D display. One such instance occurs when the mapping application of some embodiments moves the position of the virtual camera from a perspective rendering position behind the current position of a moving device to a top-down view of the current position as the device is about to turn a corner along a route. Another example is the zoom in / out animation provided that shows objects in the scene growing larger and shrinking as the zoom in and zoom out operations occur.
[0462] In addition, in some embodiments, the mapping application provides two different types of 3D presentations: an immersive 3D presentation and a non-immersive 3D presentation. The immersive presentation in some embodiments displays not only more geometry but also more detail than the geometry displayed in the non-immersive presentation. The mapping application also provides a smooth transition between the non-immersive presentation and the immersive presentation.
[0463] To achieve such smooth transitions and generate other novel effects, the mapping application of some embodiments uses a novel image processing pipeline. This pipeline performs various pre-loading operations to download, retrieve, and / or decompress map tiles that may be needed for navigation presentations, prepares its rendering pipeline for its rendering operations, and prepares a replication pipeline to smoothly transition between immersive and non-immersive 3D presentations.
[0464] Section A below describes the automatic movement of a virtual camera for rendering a 3D presentation. Section B then further describes the immersive and non-immersive 3D presentations of the mapping application. Section C then describes the image processing pipeline of the mapping application and its associated meshing pipeline, virtual camera, and rendering engine.
[0465] A. Automatic virtual camera movement
[0466] As described above, a virtual camera is a conceptualization of a location in a 3D map that is rendered by a mapping application to generate a 3D map view. A description of several different examples of automatic movement of such a virtual camera is provided below.
[0467] 1. Virtual camera movement during route navigation
[0468] Figure 74 An example of a virtual camera 7400 that automatically moves around an area of a 3D map 7435 based on route navigation is conceptually illustrated. Figure 74 The movement of the virtual camera 7400 around an area of the 3D map 7435 is shown at three different stages 7405-7415. In some embodiments, the mapping application Figure 75 To perform the process 7500 described below to render the 3D map view described in the figure. As shown in the figure, the area of 3D map 7435 includes a road extending above the bridge.
[0469] The first stage 7405 shows the virtual camera 7400 at a particular position and location in the 3D map 7435. In some embodiments, the point in the 3D map 7435 that the virtual camera 7400 is focused on (i.e., the point where a line projected from in front of the virtual camera intersects the road) is called the focus point. In some embodiments, the mapping application uses a point along the route that is a defined distance (e.g., 5 feet, 10 feet, 25 feet, 50 feet, etc.) in front of the user's location (represented in this example by a circle with a triangle on top) as the focus point for the virtual camera 7400. The mapping application of some embodiments uses different distances in front of the user's location along the route based on the user's speed of travel (e.g., a longer distance when the user is traveling faster and a shorter distance when the user is traveling slower).
[0470] In the first stage 7405, the focus point of the virtual camera 7400 is a point along the ramp of the bridge at a defined distance in front of the user's position. In this stage 7405, an exemplary rendered image of a 3D map 7435 from the perspective of the virtual camera 7400 is shown in a 3D map view 7420. The 3D map views 7420, 7425, and 7430 shown in this figure, as well as those shown in other figures in this section, do not include various UI controls and features (e.g., road signs, navigation signs, floating controls, etc.) to highlight the rendered map image.
[0471] The second stage 7410 shows the virtual camera 7400 at different positions and locations within the 3D map 7435. In this example, the virtual camera 7400 has moved upward because the focus of the virtual camera 7400 has been raised due to the height of the bridge. In addition, the virtual camera 7400 has moved to the right to navigate along the route. From the first stage 7405 to the second stage 7410, the mapping application automatically moves the virtual camera 7400, renders an image of the 3D map from the perspective of the virtual camera 7400 along the route, and displays the image on the device. In this stage 7410, an exemplary rendered image of the 3D map 7435 from the perspective of the virtual camera 7400 is shown in the 3D map view 7425.
[0472] The third stage 7415 shows the virtual camera 7400 at another different position and location within the 3D map 7435. For this example, the virtual camera 7400 has moved downward because the focus of the virtual camera 7400 is on the road and no longer on the bridge. In addition, the virtual camera 7400 has moved to the right to navigate along the route. From the second stage 7410 to the third stage 7415, the mapping application automatically moves the virtual camera 7400, renders an image of the 3D map from the perspective of the virtual camera 7400 along the route, and displays the image on the device. In this stage 7415, an exemplary rendered image of the 3D map 7435 from the perspective of the virtual camera 7400 is shown in the 3D map view 7430.
[0473] In some embodiments, the mapping application detects a threshold amount of change in elevation of a focal point along a route, or a threshold rate of change in elevation. When the mapping application detects that the user's position along the route has decreased in elevation and has passed one or both of the thresholds, the mapping application of some such embodiments adds a delay to the tracking adjusted position of the virtual camera in response to the position of the focal point along the route. In some embodiments, when the mapping application detects that the user's position along the route has decreased in elevation, the mapping application maintains the altitude of the virtual camera (e.g., the position of the virtual camera along the Z-axis) for a defined amount of time. Adding the delay and maintaining the altitude for a defined amount of time are techniques that each prevent the virtual camera from approaching too closely to the road or, in some cases, colliding with the road. In this way, the mapping application provides the user with a better map view while traversing an undulating road.
[0474] Figure 75 The process 7500 of animating a map view during route navigation in some embodiments is conceptually illustrated. The mapping application of some embodiments performs process 7500 at defined time intervals (e.g., 0.5 seconds, 1 second, 3 seconds, 5 seconds, etc.). In some embodiments, the mapping application repeatedly performs process 7500 while the mapping application operates in route navigation mode.
[0475] Process 7500 begins by identifying (at 7510) a set of properties for determining the position of a virtual camera at the end of a time period. In some embodiments, the end of a time period is the time when the mapping application receives the user's current location, and the beginning of a time period is the time when the mapping application receives the user's previous location. In some embodiments, the set of properties includes data for deriving the position and orientation of the virtual camera.
[0476] In different embodiments, process 7500 uses different sets of attributes. The following describes some or all of the attributes used by process 7500 of some embodiments. For example, process 7500 of some embodiments uses the location of a reference point (i.e., the user's location). Process 7500 of some embodiments uses the location of the focus point of the virtual camera, which, in some embodiments, is used to determine which direction the virtual camera should face.
[0477] In the event that the user deviates from the route of the turn-by-turn navigation, the focus point is a fixed distance in front of the user in the direction of travel (if the direction of travel can be determined), or a fixed distance north of the user if the direction of travel cannot be determined. In the event that the user is on the route relative to the turn-by-turn navigation, the focus point is a fixed distance in front of the user along the route to allow the virtual camera to keenly peek at turns before the user actually turns.
[0478] Some embodiments constrain the angle to be between (1) a vector defined by the reference point and the focus point, and (2) a vector pointing in the direction of travel of the user by a defined maximum angle. For example, if the route turns a corner shortly ahead, the focus point will be a point around the corner from the device's current position. Since turning the virtual camera to face a real point might cause the virtual camera to face directly into a building, the amount the virtual camera can turn to peek around a corner is limited, so that the view in the direction of travel is always partially displayed.
[0479] Another virtual camera position indicator used by process 7500 in some embodiments is the location of a point of interest (e.g., the location of an upcoming intersection). In some embodiments, process 7500 determines a virtual camera view style and uses this indicator to derive the position and orientation of the virtual camera. Examples of virtual camera view styles include top-down centered, top-down forward, and rooftop. In some embodiments, a top-down centered view style refers to a view style in which the virtual camera is looking straight down at the user's location, such that the user's location is centered on the screen. In some embodiments, a top-down forward view style refers to a view style in which the virtual camera is looking straight down at the user's location, such that the user's location is toward the bottom of the display. In some embodiments, a rooftop view style refers to a view style in which the virtual camera is positioned behind the user's location and tilted such that the virtual camera is looking forward along a vector from the user's location to the focus point. If the user deviates from the route of the turn-by-turn navigation or the user's direction of travel is uncertain (e.g., when the user is parking), the virtual camera is set to a top-down centered view style. Otherwise, the view style is determined by whether the user has requested 2D navigation mode or 3D navigation mode. If the user has requested 2D navigation mode, the view style is set to a top-down forward view style. If the user has requested 3D navigation mode, the view style is set to the roof view style.
[0480] In some embodiments, process 7500 determines a focus style for a virtual camera and uses the focus style to determine the position and orientation of the virtual camera. Examples of virtual camera focus styles include a cruising focus style and a hard focus style. In some embodiments, a cruising focus style refers to a style in which the virtual camera is set to a preset altitude and pitch angle. In some embodiments, a hard focus style refers to a style in which the virtual camera is set to an altitude (e.g., in a top-down centered or top-down forward view style) and / or a pitch (e.g., in a rooftop view style) such that a given point of interest is displayed on the screen (e.g., the virtual camera focuses on the point of interest as the user approaches the point of interest).
[0481] Next, the process 7500 determines (at 7520) the position and orientation of the virtual camera based on the identified attributes. After determining the position and orientation of the virtual camera, the process 7500 identifies (at 7530) the position of a set of reference points at the end of the time period based on the position and orientation of the virtual camera. Examples of reference points include the position of the camera, the intersection between the forward vector of the virtual camera and the ground, a defined point along the right vector of the virtual camera, and the like.
[0482] Process 7500 then identifies (at 7540) the locations of the corresponding set of previous reference points. In some embodiments, the mapping application maintains a history of the locations of the set of reference points, determined in the same manner as those identified by process 7500 at 7520, and process 7500 accesses this history.
[0483] Finally, the process 7500 renders (at 7550) an animated map view for the time period based on the set of reference points identified at 7530 and 7540. To render the animated map view, the process 7500 of some embodiments uses a cubic polynomial associated with each reference point. The process 7500 animates the map view by (1) traversing each of the virtual camera's reference points along the reference point's associated cubic polynomial during the animation, while simultaneously (2) rendering the virtual camera's map view.
[0484] Before rendering an animated map view, when a reference point has changed relative to its corresponding previous reference point, the process 7500 of some embodiments updates a cubic polynomial associated with a particular reference point. In some embodiments, the process 7500 updates the cubic polynomial using a cubic polynomial determined based on the position of the reference point, the position of the corresponding previous reference point, and two tangents based on the reference point and its corresponding previous reference point. In some embodiments, the process 7500 uses a scaled version of a vector that ends at the position of the reference point and starts at the position of the corresponding previous reference point as one of the tangents. For the other tangent, the process 7500 of some embodiments uses another scaled vector determined in a similar manner to the first scaled vector, except that the vector ends at the position of the previous reference point and the position of its corresponding previous reference point. The process 7500 fits a cubic polynomial using the two tangents, the position of the reference point, and the position of the corresponding previous reference point, and uses the fitted cubic polynomial as the current cubic polynomial for the reference point.
[0485] 2. Virtual camera movement around the intersection
[0486] Figure 76 An example of a virtual camera 7600 automatically maneuvering turns in a 3D map 7650 during route navigation is conceptually illustrated. Figure 76The movement of the virtual camera 7600 in the 3D map 7650 is shown in three different stages 7605-7615. In some embodiments, the mapping application Figure 77 To perform the process 7700 described below to manipulate the turn. As shown, the area of the 3D map 7650 includes a road adjacent to another road and a building on the right side of the T-junction.
[0487] Stage 1 7605 shows a virtual camera 7600 at a particular position and location in a 3D map 7650. In some embodiments, the point in the 3D map 7650 that the virtual camera 7600 is focused on (i.e., the point where a line projected from in front of the virtual camera intersects the road) is called the focus point. In some embodiments, the mapping application uses a point along the route that is a defined distance (e.g., 5 feet, 10 feet, 25 feet, 50 feet, etc.) in front of the user's location (represented in this example by a circle with a triangle on top) as the focus point for the virtual camera 7600. The mapping application of some embodiments uses different distances in front of the user's location along the route based on the user's travel speed (e.g., a longer distance when the user is traveling faster and a shorter distance when the user is traveling slower). For this example, the focus point of the virtual camera 7600 is the point along the route in front of the user's location.
[0488] In the first stage 7605, the focus point of the virtual camera 7600 is a point in front of the user's position on the road along which the user is traveling. The first stage 7605 also shows a corresponding top-down view 7620 of the virtual camera 7600 and the 3D map 7650. As shown, in this example, the focus point is indicated by a black circle located in front of the user on the road along which the user is traveling. In this stage 7605, an exemplary rendered image of the 3D map 7650 from the perspective of the virtual camera 7600 is shown in the 3D map view 7635. The 3D map view 7635 also shows the focus point of the virtual camera.
[0489] The second stage 7610 shows the virtual camera 7600 at different positions and locations within the 3D map 7650. As shown in the second stage 7610, the virtual camera 7600 has moved to the right along the route navigation toward the intersection. To maintain focus on the focal point, the virtual camera 7600 has also moved upward (e.g., upward along the Z axis) to provide a better view of the focal point during the right turn in case the virtual camera is obstructed. For example, when there are buildings nearby or at the intersection, the focal point may be obstructed by the buildings, thereby obstructing the view ahead of the turn corner.
[0490] In addition, the virtual camera 7600 has been tilted downward (e.g., reduced pitch) at this stage to maintain aiming at the focus point. The top-down view 7625 at the second stage 7610 shows that the virtual camera 7600 has been rotated slightly to the right side of the intersection (e.g., about the Z axis) because the focus point is just toward the right side of the intersection. Thus, from the first stage 7605 to the second stage 7610, the mapping application has automatically moved the virtual camera 7600 forward along the route navigation (e.g., along the X axis, Y axis, or a combination of both the X axis and the Y axis, etc.) toward the intersection, moved the virtual camera 7600 higher (e.g., upward along the Z axis), and rotated the virtual camera 7600 toward the right side of the intersection (e.g., about the Z axis) while maintaining focus on the focus point in front of the user's position. In the second stage 7610, an exemplary rendered image of the 3D map 7650 from the perspective of the virtual camera 7600 is shown in the 3D map view 7640.
[0491] The third stage 7615 shows the virtual camera 7600 at another different position and location in the 3D map 7650. Here, the virtual camera 7600 has moved to the right and is at an intersection along the route navigation. To maintain focus on the focal point, the virtual camera 7600 has also moved upward (e.g., upward along the Z axis) and above the intersection to provide a better view of the focal point while the user is turning right for the same or similar reasons described above (e.g., an obstruction or occlusion of the view of the focal point of the virtual camera 7600).
[0492] In the third stage 7615, the virtual camera 7600 has been tilted further downward (e.g., reduced pitch) to maintain focus on the focal point. The top-down view 7630 in the third stage 7615 shows that the virtual camera 7600 has been rotated to the right (e.g., about the Z axis) because the focal point has now turned to the right turn of the intersection. Thus, from the second stage 7610 to the third stage 7615, the mapping application has automatically moved the virtual camera 7600 forward along the route navigation (e.g., along the X axis, the Y axis, or a combination of the X axis and the Y axis, etc.) to the intersection, moved the virtual camera 7600 higher (e.g., upward along the Z axis), and rotated the virtual camera 7600 toward the right side of the intersection (e.g., about the Z axis) while maintaining focus on the focal point in front of the user's position. In the third stage 7615, an exemplary rendered image of the 3D map 7650 from the perspective of the virtual camera 7600 is shown in the 3D map view 7645.
[0493] For purposes of illustration and clarity, Figure 76Shown is a focus point in a top-down view of a 3D map and a rendered 3D map view of the 3D map.In some embodiments, the mapping application does not actually render the focus point in the 3D map view, and the 3D map does not contain an actual point representing the focus point.
[0494] Figure 77 Conceptually illustrated is a process 7700 of some embodiments for animating a map view during turns in a route navigation. In some embodiments, the mapping application iteratively performs the process 7700 as the mapping application iterates through the route navigation. The mapping application of some embodiments performs the process 7700 at defined time intervals (e.g., 0.5 seconds, 1 second, 3 seconds, etc.) as the mapping application iterates through the route navigation.
[0495] Process 7700 begins by identifying (at 7710) an upcoming intersection. In some embodiments, process 7700 uses map data to identify an upcoming intersection along a route to navigate.
[0496] Next, the process 7700 determines (at 7720) whether the navigation route has a turn at the upcoming intersection. When the process 7700 of some embodiments identifies a threshold change near or at the intersection or determines that the street name at the intersection along the navigation route changes, the process 7700 determines that the navigation route has a turn at the upcoming intersection. Additional techniques and / or other techniques are possible.
[0497] When process 7700 determines that there is no turn at the next intersection, process 7700 ends. Otherwise, process 7700 determines (at 7730) whether the turn angle is within a threshold angle (e.g., whether the curve deviates from a straight line by more than a certain number of degrees). The threshold angle is defined differently for different embodiments. In some embodiments, the threshold angle is defined as a value less than 45 degrees, a value between 45 and 90 degrees, a value greater than 90 degrees, etc.
[0498] If process 7700 determines that the turn angle is within the threshold angle, process 7700 ends. Otherwise, process 7700 displays (at 7740) the map view in a top-down viewing mode. In some embodiments, top-down mode is a viewing mode in which the virtual camera is located above or "on top" of the scene and is pointing mostly or entirely downward. In some embodiments, the top-down viewing mode is referred to as a hard focus mode. In some embodiments, the top-down viewing mode is a 2D mode. In some cases, the mapping application is already displaying a map view in top-down mode. In such cases, process 7700 continues in this viewing mode by navigating turns along the route.
[0499] Next, process 7700 determines (at 7750) whether the turn has been completed. When the user's position is a defined distance beyond the intersection, process 7700 determines that the turn has been completed. In some embodiments, process 7700 determines that the turn has been completed when the virtual camera's position has rotated by an amount equal to or within a threshold angle range of the turn angle. If process 7700 determines that the turn has not been completed, process 7700 returns to 7750 to continue checking whether the turn has been completed.
[0500] When process 7700 determines that the turn has been completed, process 7700 displays (at 7760) the map view in the original viewing mode. That is, process 7700 displays the map view in the viewing mode in which the map view was being displayed before mapping was applied to the approaching turn. Process 7700 then ends.
[0501] Although process 7700 is described as using a top-down mode, in some embodiments, other similar or identical processes use a non-top-down mode that makes turns around corners easier to see than in any mode in which the map is displayed before approaching the corner.
[0502] 3. Virtual camera movement during panning
[0503] Figure 78 An example of a virtual camera 7800 panning around a 3D map 7835 during route navigation is conceptually illustrated. Figure 78 The movement of the virtual camera 7800 around an area of the 3D map 7835 is shown in three different stages 7805-7815. The mapping application of some embodiments is described by referring to Figure 79 The process 7900 described below is performed to translate the virtual camera 7800. As shown in the figure, the area of the 3D map 7835 includes a road and several buildings along the side of the road.
[0504] The first stage 7805 shows a top-down view of a virtual camera 7800 at a particular position and location in a 3D map 7835. In some embodiments, the point in the 3D map 7835 that the virtual camera 7800 is focusing on (i.e., the point where a line projected from in front of the virtual camera intersects the road) is called the focus point. In some embodiments, the mapping application uses a point along the route that is a defined distance (e.g., 5 feet, 10 feet, 25 feet, 50 feet, etc.) in front of the user's location (represented in this example by a circle with a triangle on top) as the focus point for the virtual camera 7800. The mapping application of some embodiments uses different distances in front of the user's location along the route based on the user's travel speed (e.g., a longer distance when the user is traveling faster and a shorter distance when the user is traveling slower).
[0505] In the first stage 7805, the focus point of the virtual camera 7800 is a point along the road a defined distance ahead of the user's location, as indicated by a circle in the 3D map 7835. In the first stage 7805, an exemplary rendered image of the 3D map 7835 from the perspective of the virtual camera 7800 is shown in the 3D map view 7820.
[0506] The second stage 7810 shows a top-down view of the virtual camera 7800 at different positions and locations in the 3D map 7835. As shown, the virtual camera 7800 has been rotated to the right (e.g., about the Z axis). In this example, the mapping application of some embodiments automatically rotates the virtual camera 7800 to the right in response to gesture input received via the touch screen (e.g., a left swipe gesture). In this stage 7810, an exemplary rendered image of the 3D map 7835 from the perspective of the virtual camera 7800 is shown in the 3D map view 7825.
[0507] The third stage 7815 shows a top-down view of the virtual camera 7800 at another different position and location of the 3D map 7835. For this example, the mapping application has automatically rotated the virtual camera 7800 back to the left after a defined amount of time (e.g., 0.5 seconds, 1 second, 2 seconds, 3 seconds, etc.). In some embodiments, the mapping application rotates the virtual camera 7800 back after traveling a defined distance (e.g., 10 yards, 25 yards, half a mile, etc.). In the third stage 7815, an exemplary rendered image of the 3D map 7835 from the perspective of the virtual camera 7800 is shown in the 3D map view 7830.
[0508] Figure 79 Conceptually illustrating some embodiments of process 7900 for panning to the side of a map view based on gesture input, in some embodiments, the mapping application performs process 7900 when the mapping application is in turn-by-turn navigation mode and the mapping application receives a gesture via a touchscreen of the device on which the mapping application is executing.
[0509] Process 7900 begins by receiving (at 7910) a gesture for panning to the side. In some embodiments, the gesture for panning to the side comprises a touch gesture received via the touch screen (e.g., swiping the touch screen to the side with a finger). Alternatively or in combination, in some embodiments, the gesture for panning to the side comprises a multi-touch gesture (e.g., swiping the touch screen to the side with multiple fingers).
[0510] Next, the process 7900 identifies (at 7920) a translation component of the pan gesture. The process 7900 of some embodiments identifies the horizontal component of the pan gesture as the translation component. In other embodiments, the process 7900 identifies a vector defined from the location of the initial touch of the gesture (e.g., where the finger touched the screen) to the location where the gesture ended (e.g., where the finger no longer touched the touch screen).
[0511] In some embodiments, process 7900 limits the translation component of the pan gesture. For example, if process 7900 determines (at 7930) that the translation component exceeds a threshold amount (e.g., the magnitude of the translation component exceeds the threshold amount), process 7900 limits (at 7940) the translation component (e.g., reduces the magnitude of the translation component) to within a defined limit.
[0512] Based on the translation component, the process 7900 then translates (at 7950) the virtual camera being used to render the 3D map view of the route for route navigation. The process 7900 of some embodiments is similar to that described above with reference to FIG. Figure 78 The process 7900 translates the virtual camera in a similar manner to the process described above, that is, the process 7900 rotates the virtual camera around the Z axis to translate the virtual camera.
[0513] Next, the process 7900 determines (at 7960) whether a defined amount of time has elapsed (e.g., 0.5 seconds, 1 second, 2 seconds, 5 seconds, etc.). If the process 7900 determines that the defined amount of time has not elapsed, the process 7900 returns to 7960 to continue checking for an amount of time that will elapse. If the process 7900 determines that the defined amount of time has elapsed, the process 7900 proceeds to 7970.
[0514] Finally, the process 7900 translates (at 7970) the virtual camera back to its original position. In some embodiments, the process 7900 is performed in the same manner as described above with reference to FIG. Figure 78 The process 7900 translates the virtual camera in a similar manner to the process 7900 described above. In other words, the process 7900 rotates the virtual camera around the Z axis to translate the virtual camera.
[0515] 4. Virtual Camera Movement During View Adjustment
[0516] In addition to (or instead of) having the navigation application control the camera (e.g., switching from 3D to 2D when walking around a corner), some embodiments allow the user to adjust the camera's level. Some embodiments allow the user to make command gestures using two fingers to adjust the camera's distance (height) and angle. Some embodiments even allow multiple types of gestures to control the camera.
[0517] Figure 80 The perspective adjustment feature provided by the mapping application of some embodiments is conceptually illustrated. Specifically, Figure 80The virtual camera 8000 is shown in three different stages 8005-8015, which illustrate adjustments to the position of the virtual camera 8000 in response to viewing angle adjustments. Figure 80 A location in a 3D map 8035 is shown that includes four objects, namely two buildings and two intersecting roads.
[0518] The first stage 8005 shows a virtual camera 8000 at a first perspective position, which is pointing downward at a 3D map 8035 at a first angle (e.g., 45 degrees) relative to the horizontal. In this position, the camera 8000 is pointing to a location that can be the resting position of the device or the resting position of the location being explored, or a moving position that is ahead of the moving position of the device when the map is used for navigation. In some embodiments, the default position of the camera 8000 will be in a specific orientation relative to the current position, but this orientation can be modified when the user rotates the map. Rendering a 3D map view based on the position of the virtual camera 8000 results in a 3D map view 8025.
[0519] The second stage 8010 shows the virtual camera 8000 at a different, second perspective position, pointing at the 3D map 8035 at a lower perspective at a smaller second angle (e.g., 30 degrees) relative to the horizontal. Stage 8010 also shows that the user has provided input to adjust the perspective of the view of the 3D map 8035 by touching the screen with two fingers and dragging them in an upward direction (e.g., a swipe gesture). The scene tiling is achieved by lowering the virtual camera 8000 and reducing the viewing angle relative to the horizontal. Rendering the 3D map view using the virtual camera 8000 at this angular position results in a 3D map view 8030 in which buildings and roads are taller than they were in the first map view 8025. As indicated by the dashed version of the virtual camera 8000, the virtual camera 8000 moves further downward along an arc 8050 while tilting further upward (e.g., pitching).
[0520] The third stage 8015 shows the virtual camera 8000 at a different third perspective position, with the virtual camera pointing at a location on the 3D map 8035 (e.g., the focus point of the virtual camera 8000) at a higher perspective at a larger third angle (e.g., 80°) relative to the horizontal. Stage 8015 also shows that the user has provided input to adjust the perspective of the view of the 3D map 8035 by touching the screen with two fingers and dragging them in a downward direction (e.g., a swipe gesture). The scene descent is achieved by raising the virtual camera 8000 and increasing its angle relative to the horizontal. As shown, at this stage 8015, in some embodiments, when the virtual camera 8000 is in a top-down or nearly top-down position, the mapping application tiles the buildings in the 3D map 8035 (e.g., lowering the Z-axis component of the polygons to ground level), so that the 3D map view rendered using the virtual camera 8000 appears to be 2D. Rendering the 3D map view in the third stage 8015 using the virtual camera 8000 at this angular position results in a 3D map view 8040 in which buildings appear smaller and flatter, and roads appear smaller, than they were in the second map view 8030. As indicated by the dashed version of the virtual camera 8000, the virtual camera 8000 moves further upward along the arc 8050 while being tilted further downward (e.g., pitched).
[0521] In some embodiments, the virtual camera 8000 may be caused to move in this manner when the mapping application receives input for adjusting the viewing angle for viewing the 3D map 8035. In some such embodiments, when the zoom level reaches a particular zoom-out level, the mapping application switches to a top-down mode (with the rendered location facing directly downward) that produces a 2D map view.
[0522] Figure 80 The virtual camera is shown moving along a specific path along an arc. The mapping application of different embodiments moves the virtual camera along different paths when adjusting the viewing angle of the virtual camera. For example, in some embodiments, the mapping application uses Figure 80 , except that as the virtual camera moves up the arc and moves closer (e.g., zooming in), the particular arc is further away and upward (e.g., zooming out), and as the virtual camera moves down the arc, the particular arc is downward. In some embodiments, the mapping application uses a path for the virtual camera that tiles horizontally as the virtual camera moves up the arc. In other embodiments, other paths along which the virtual camera moves are possible.
[0523] When moving along one of the arcs, the virtual camera rotates to maintain the desired location on the map at the camera's focus. In some cases, the desired location is the device's static location or the static location the user is browsing on the map. In other cases, the desired location is a moving location that is ahead of the device's moving location as the user is moving with the device.
[0524] Figure 81 A process 8100 of some embodiments is conceptually illustrated for adjusting the position of a virtual camera for rendering a map view in response to a viewing angle adjustment. In some embodiments, when a mapping application is in a map viewing mode (e.g., a location browsing mode, a navigation mode, etc.) and the mapping application receives a gesture via a touch screen of a device on which the mapping application is executing, the above and below references Figure 1-108 The described mapping application performs process 8100.
[0525] Process 8100 begins by receiving (at 8110) a gesture for adjusting the perspective of a map view. In some embodiments, the gesture for adjusting the perspective of a map view comprises a multi-touch gesture (e.g., touching the touch screen with multiple fingers simultaneously) received via the touch screen, such as described above with reference to FIG. Figure 80 In this example, process 8100 receives a dual-touch swipe gesture.
[0526] Next, process 8100 determines (at 8120) whether the angle based on the touch point is within a threshold angle. In some embodiments, the angle is defined by: (1) a first vector from the location of one of the touch points toward the location of the other touch point, and (2) a second vector from the location of the one touch point toward a horizontal direction. The horizontal direction for some embodiments is based on the orientation of the device, while the horizontal direction for other embodiments is based on the orientation of the device's display.
[0527] When process 8100 determines that the angle is within the threshold angle, process 8100 ends. Otherwise, process 8100 identifies (at 8130) the translational component of the received gesture. In some embodiments, process 8100 identifies the translational component of the gesture by identifying the amount of vertical translation of the touch point of the gesture.
[0528] The process 8100 then adjusts (at 8140) the perspective of the map view based on the identified translation. Figure 3-5 and Figures 74-80As described above, in some embodiments, a virtual camera is used to render a map view. Process 8100 of some embodiments adjusts the perspective of the map view by adjusting the position and orientation of the virtual camera. In some embodiments, a downward vertical translation causes the position of the virtual camera to increase in altitude and decrease in pitch (e.g., point further downward) while maintaining the focus point of the virtual camera (e.g., the point where the forward vector of the virtual camera intersects the ground). In other words, the camera moves along an arc and rotates downward more as the camera moves upward along the arc.
[0529] Different embodiments use different coordinate spaces for maps. For example, some embodiments use a Mercator unit coordinate space for their maps. In such embodiments, process 8100 adjusts the position of the virtual camera relative to the map to adjust the perspective of the map view. For another example, in some embodiments, the map uses the World Geodetic System (e.g., WGS 84) as its coordinate space. In such embodiments, process 8100 adjusts the map relative to the position of the virtual camera to adjust the perspective of the map view.
[0530] In some embodiments, an upward vertical translation causes the position of the virtual camera to decrease in altitude and increase in pitch (e.g., point further upward) while maintaining the virtual camera's focal point (e.g., the point where the virtual camera's forward vector intersects the ground). That is, the camera moves along an arc and rotates upward more as the camera moves downward along the arc. In different embodiments, the vertical translation causes different adjustments to the position of the virtual camera. For example, a downward vertical translation may cause the position of the virtual camera to decrease in altitude and increase in pitch while maintaining the virtual camera's focal point, and an upward vertical translation may cause the position of the virtual camera to increase in altitude and decrease in pitch while maintaining the virtual camera's focal point.
[0531] Finally, process 8100 renders (at 8150) the map view for display based on the adjusted perspective of the map view. Process 8100 then ends.
[0532] As mentioned above, Figure 81The illustrated process illustrates adjusting the perspective of a map view based on gesture input. The process of some embodiments limits the range of movement of the virtual camera. When the gesture input causes the position of the virtual camera to exceed a defined range, the process of some embodiments modifies the position of the virtual camera to be within the defined range. In some such embodiments, even if the position of the virtual camera exceeds the defined range during the gesture input (e.g., the finger used for the touch point is still touching the touch screen), the process renders a map view of the virtual camera's position. When the gesture input is completed (e.g., the finger used for the touch point is no longer touching the touch screen), the process renders a map view of the virtual camera's position that is modified to be within the defined range. In some embodiments, when the gesture input is completed, the process displays an animation that "jumps" from the map view of the virtual camera's position that is outside the defined range to the map view of the virtual camera's position that is modified to be within the defined range. In some embodiments, such animation and the limiting feature may be referred to as an offset rubber band feature or effect.
[0533] like Figure 80 As shown, different virtual camera positions render different perspectives of the map view. Specifically, Figure 80 8005 , the buildings will appear to be rising from the ground. In some embodiments, the mapping application scales the vertices of the buildings along the Z axis so that the buildings appear to rise and fall as the map view transitions between the different perspectives of the 2D-like map view and the 3D map view. The following is an example of an equation used by the mapping application of some embodiments to scale the vertices along the Z axis:
[0534] new_scaled_vertex.z=scaled_vertex.z*u_scale
[0535] where scaled_vertex.z is the Z component of the vertex, u_scale is a scaling factor that determines the amount to scale, and new_scaled_vertex.z is the new scaled value of the vertex's Z component. Additionally, some embodiments interpolate alpha values for blending in and out of buildings and interpolate scale values for adjusting the height of buildings when zooming in / out or moving around the 3D map.
[0536] As mentioned above, Figure 80Several exemplary perspective adjustment operations and corresponding movements of a virtual camera in a 3D map are shown, the corresponding movements being used to render a 3D map view of the 3D map. Those skilled in the art will appreciate that any number of different perspective adjustments are possible. Furthermore, the mapping application of some embodiments performs perspective adjustment operations in response to additional and / or different types of input. For example, in some embodiments, dragging two fingers downward may lower the perspective position of the virtual camera, and dragging two fingers upward may increase the perspective position of the virtual camera.
[0537] Although Figure 80 The virtual camera is shown moving along an arc to change the viewing position of the virtual camera, but the mapping application of some embodiments moves the virtual camera along different arcs at different zoom levels (e.g., distance from the 3D map). Figure 82 A virtual camera 8210 is conceptually shown that can be moved along three different arcs 8215-8225 at three different zoom levels. As shown, Figure 82 A location in a 3D map 8205 is shown, which includes two buildings and two roads forming a T-junction.
[0538] At zoom level 18 (the lowest zoom level in this example), the virtual camera 8210 can move along an arc 8215 for different viewing angle positions of the virtual camera 8210 directed toward the 3D map 8205. In other words, the mapping application of some embodiments at this zoom level allows the user to adjust the viewing angle for viewing the 3D map 8205 to within the range (e.g., a 20-degree angular range) within which the virtual camera 8210 can move along the arc 8215.
[0539] When the virtual camera 8210 transitions from zoom level 19 (e.g., zooms out), the mapping application limits the range of the viewing angle position of the virtual camera 8200 to a range along arc 8215. In some embodiments, the mapping application positions the virtual camera 8210 at a position along arc 8215 that corresponds to the position of the virtual camera 8210 along arc 8220 at zoom level 19 prior to the transition to zoom level 18. If the position of the virtual camera 8210 at zoom level 19 is outside the range of arc 8215 (e.g., a position along the lower viewing angle end of arc 8220) when transitioning from zoom level 19 to zoom level 18, in some embodiments, the mapping application positions the virtual camera 8210 as close as possible. In other embodiments, the mapping application gradually moves the virtual camera 8210 toward the nearest corresponding position on the arc 8215 while being in a transition region between zoom levels 18 and 19 when transitioning from zoom level 19 to zoom level 18 to provide a smooth transition of the movement of the virtual camera 8210 from zoom level 19 to zoom level 18.
[0540] At the next zoom level, 19, the virtual camera 8210 can move along an arc 8220 for different viewing angle positions of the virtual camera 8210 directed toward the 3D map 8205. As shown, in some embodiments, the mapping application at this zoom level allows the user to adjust the viewing angle for viewing the 3D map 8205 to within the range (e.g., a 45-degree angular range) within which the virtual camera 8210 can move along the arc 8220. Thus, at zoom level 19, the mapping application allows the virtual camera 8210 to move over a greater range than at zoom level 18.
[0541] When the virtual camera 8210 transitions from zoom level 18 or zoom level 20 (e.g., zooms in or out), the mapping application limits the range of the view angle position of the virtual camera 8200 to a range along arc 8220. The mapping application of some embodiments positions the virtual camera 8210 at a position along arc 8220 that corresponds to the position of the virtual camera 8210 along arc 8215 or 8225 at zoom level 18 or 20, respectively, before transitioning to zoom level 19. If the position of the virtual camera 8210 at zoom level 20 is outside the range of arc 8220 (e.g., along the lower view angle end of arc 8225) when transitioning from zoom level 20 to zoom level 19, the mapping application, in some embodiments, positions the virtual camera at a position as close as possible. The mapping application of other embodiments gradually moves the virtual camera 8210 toward the nearest corresponding position on the arc 8220, while being in the transition area between zoom levels 19 and 20 when transitioning from zoom level 20 to zoom level 19, to provide a smooth transition of the movement of the virtual camera 8210 from zoom level 20 to zoom level 19.
[0542] At zoom level 20 (the highest zoom level in this example), the virtual camera 8210 can move along arc 8225 for different viewpoint positions of the virtual camera 8210 directed toward the 3D map 8205. That is, in some embodiments, the mapping application at this zoom level allows the user to adjust the viewpoint for viewing the 3D map 8205 to within the range (e.g., an 85-degree angular range) within which the virtual camera 8210 can move along arc 8225. Thus, at zoom level 20, the mapping application allows the virtual camera 8210 to move along a greater range than at zoom levels 18 and 19. When the virtual camera 8210 transitions from zoom level 19 (e.g., zooms in), the mapping application limits the range of viewpoint positions of the virtual camera 8200 to the range along arc 8225.
[0543] above Figure 82An example of a virtual camera that can be moved along different arcs at different zoom levels is shown, wherein the different zoom levels allow for different ranges of viewing angle positions for the virtual camera. One of ordinary skill in the art will recognize that mapping applications of different embodiments use any number of different arcs and ranges of viewing angle positions for the virtual camera at different zoom levels.
[0544] Figure 83 The process 8300 of some embodiments for determining an arc along which a virtual camera can move is conceptually illustrated. In some embodiments, when adjusting the zoom level (e.g., zooming in or out) in a 3D viewing mode (e.g., 3D navigation mode, 3D map browsing mode, etc.), the above and below references in this application Figure 1-108 The described mapping application performs process 8300.
[0545] Process 8300 begins by determining (at 8310) whether to adjust the current zoom level to a new zoom level. In some embodiments, process 8300 receives input for adjusting the zoom level of the view of the 3D map (e.g., as described below with reference to FIG. Figure 85 When the input type is selected, process 8300 determines to adjust the current zoom level to the new zoom level. When process 8300 determines not to adjust the current zoom level, process 8300 returns to 8310 to continue checking for adjustment to the new zoom level.
[0546] When process 8300 determines that an adjustment for a new zoom level has been received, process 8300 identifies (at 8320) an arc for the new zoom level. As described above, some embodiments of the mapping application use different arcs at different zoom levels. Thus, process 8300 identifies the arc defined for the new zoom level.
[0547] Next, the process 8300 determines (at 8330) whether the arc of the new zoom level has a position that corresponds to the position of the virtual camera along the previous arc at the previous zoom level. When the position of the virtual camera along the previous arc is within the angular range of the arc of the new zoom level, the process 8300 of some embodiments determines that the arc of the new zoom level has a corresponding position. Figure 82 As an example, if the zoom level is transitioning from zoom level 19 to zoom level 18, when the position of the virtual camera along arc 8220 for zoom level 19 is within approximately the upper third of arc 8220, process 8300 determines that arc 8215 for zoom level 18 has a corresponding position.
[0548] When it is determined that the arc of the new zoom level has a corresponding position, the process 8300 identifies (at 8340) the corresponding position along the arc of the new zoom level as the new position of the virtual camera. Otherwise, the process 8300 identifies (at 8350) the position along the arc of the new zoom level that is closest to the position of the virtual camera along the previous arc of the previous zoom level as the new position of the virtual camera. Figure 82 For example, if the virtual camera is positioned at the lower viewing angle end of arc 8220, as shown in FIG. Figure 82 As shown, when transitioning from zoom level 19 to zoom level 18, process 8300 will Figure 82 The position of the virtual camera shown along arc 8215 at zoom level 18 is identified as the new position of the virtual camera.
[0549] Finally, process 8300 positions the virtual camera (at 8360) at the identified new position along the arc at the new zoom level. Process 8300 then ends.
[0550] Although Figure 82 and Figure 83 While a virtual camera is described that can move along different arcs at different zoom levels, some embodiments use a specific arc for a range of different zoom levels (e.g., an arc for zoom levels 10-13, an arc for zoom levels 14-16, etc.) In some embodiments, different ranges of zoom levels are defined to include the same or different numbers of zoom levels.
[0551] As discussed above, in some embodiments, the mapping application uses a specific arc along which the virtual camera can move within a specific defined angle range along the arc when the virtual camera is positioned at a specific zoom level. The mapping application of some embodiments uses the following techniques to move the virtual camera along the arc.
[0552] Figure 84 Conceptually illustrates a feature provided by a mapping application of some embodiments for maintaining a virtual camera position within a defined range along an arc. Specifically, Figure 84 The virtual camera 8400 is shown at three different stages 8405-8415 showing the position of the virtual camera 8400 maintained within the definition of the arc 8450. Figure 84 As shown, the location in the 3D map 8435 includes two buildings and two roads forming a T-junction.
[0553] The first stage 8405 shows the virtual camera 8400 at a particular position along an arc 8450. As shown, the arc 8450 represents a defined range (e.g., an angular range) within which the virtual camera 8400 can move. The first stage 8405 also shows three positions 8455-8465 (e.g., viewing angles) along the arc 8450. In this example, the mapping application is similar to the one described above with reference to FIG. Figure 80 The described manner moves virtual camera 8400 along arc 8450 between the high-view end of arc 8450 (e.g., the position along arc 8450 when virtual camera 8400 is tilted most downward) and position 8455. In the first stage 8405, rendering a 3D map view based on the position of virtual camera 8400 results in 3D map view 8425.
[0554] As the virtual camera 8400 passes through position 8455 while moving toward the low-angle end of the arc 8450, the mapping application reduces the speed (e.g., decelerates) of the virtual camera 8400 moving toward the low-angle end of the arc 8450, regardless of the input provided by the user. In some embodiments, the mapping application reduces the speed of the virtual camera 8400 at a constant rate, while in some embodiments, the mapping application reduces the speed of the virtual camera 8400 at an exponential rate. In some embodiments, additional and / or different methods of reducing the speed of the virtual camera 8400 are used.
[0555] The second stage 8410 shows that the virtual camera 8400 has moved to a position along the arc 8450 at or near the low angle end of the arc 8450. As shown, the user is providing input to adjust the angle of view of the 3D map 8435 by touching two fingers to the screen and dragging the two fingers in an upward direction (e.g., a swipe gesture). In response to the input, the mapping application moves the virtual camera 8400 toward the low angle end of the arc 8450 while tilting the virtual camera 8450 upward. When the virtual camera reaches position 8465 along the arc 8450, the mapping application prevents the virtual camera 8400 from moving below position 8465 while the user continues to provide input for lowering the angle of view of the 3D map 8435 (e.g., the user continues to drag the two fingers upward on the touch screen).
[0556] In some embodiments, when the user stops providing input for lowering the perspective of the view of the 3D map 8435 (e.g., the user lifts two fingers off the touch screen), the mapping application causes the position of the virtual camera 8400 to "bounce" or "jump" upward from position 8465 to position 8460 along arc 8450. Because the mapping application is generating or rendering a 3D map view of the 3D map 8435 based on the view of the virtual camera 8400 during the bounce or jump motion, the generated 3D map view provides a bounce animation showing the 3D map view quickly bouncing or jumping downward to indicate to the user that the perspective of the map view cannot be lowered further. Rendering the 3D map view using the virtual camera 8400 at this angular position results in a 3D map view 8430 in which buildings and roads are taller than in map view 8425.
[0557] The third stage 8415 shows the virtual camera 8400 after the mapping application rebounds or jumps the position of the virtual camera 8400 to position 8460 in response to the user ceasing to provide input. Different embodiments use different techniques to implement the rebound or jump of the virtual camera 8400. For example, the mapping application of some embodiments begins to rapidly accelerate the virtual camera 8400 along the arc 8450 for a defined distance or before the virtual camera 8400 reaches a defined speed. The mapping application then decelerates the virtual camera 8400 for the remaining distance along the arc 8450 to the position 8460. In some embodiments, other methods are used to implement the rebound or jump effect. In the third stage 8415, rendering the 3D map view using the virtual camera 8400 at position 8460 along the arc 8450 results in a 3D map view 8440 in which the buildings appear smaller, flatter, and the roads appear smaller than in the map view 8430.
[0558] As mentioned above, Figure 84 Techniques for preventing the virtual camera from moving beyond the low-angle end of the arc are shown. Alternatively or in combination with preventing the virtual camera from moving beyond the low-angle end of the arc, the mapping application of some embodiments uses similar techniques to prevent the virtual camera from moving beyond the high-angle end of the arc. In addition, Figure 84 Examples of the following positions are shown: a position along the arc at which the virtual camera is slowed, a position along the arc to prevent the virtual camera from moving beyond, and a position along the arc to which the virtual camera jumps or rebounds. Different embodiments define the positions in any number of different ways. For example, in some embodiments, the position along the arc at which the virtual camera is slowed is the same as or near the position along the arc to which the virtual camera jumps or rebounds.
[0559] 5. Virtual camera movement during zoom
[0560] Alternatively or in combination with the viewing angle adjustment feature, the mapping application of some embodiments allows the user to zoom in and out of the view of the 3D map (e.g., by providing gesture input with two fingers). Figure 85 The zoom adjustment feature provided by the mapping application of some embodiments is shown. Specifically, Figure 85 The virtual camera 8512 is shown in three different stages 8501-8503, which show the movement of the virtual camera 8512 in response to the zoom adjustment. Figure 85 The location in the 3D map 8510 is shown to include two buildings and two roads forming a T-junction.
[0561] In the first stage 8501, a 3D map 8510 is displayed with a virtual camera 8512 at a specific location pointing at the 3D map 8510. In this location, the camera 8512 is pointed at a location, which can be the static location of the device or the location being explored, or a moving location ahead of the device's moving location when the map is used for navigation. Rendering a 3D map view based on the position of the virtual camera 8512 results in a 3D map view 8514.
[0562] The second stage 8502 shows a virtual camera 8512 at different zoom level positions pointing at the 3D map 8510. Stage 8502 shows that the user has provided input to increase the zoom level of the view of the 3D map 8510 by touching the screen of the device with two fingers close to each other and moving the fingers apart while the fingers are touching the screen (e.g., a spread gesture).
[0563] The zoom adjustment is achieved by moving the virtual camera 8512 closer to the 3D map 8510 along a line 8550. In some embodiments, the line 8550 along which the mapping application moves the virtual camera 8512 is a line formed by the front of the virtual camera 8512 and the focal point of the virtual camera 8512. The mapping application of some embodiments moves the virtual camera 8512 along the line formed by the front of the virtual camera 8512 and a location in the 3D map 8510 based on user input to zoom in on the view of the 3D map 8510.
[0564] Rendering a 3D map view using virtual camera 8512 at this location results in 3D map view 8524 in which buildings and roads appear closer than shown in 3D map view 8514. As indicated by the dashed version of virtual camera 8512, virtual camera 8512 moves closer to 3D map 8510 along line 8550.
[0565] The third stage 8503 shows a virtual camera 8512 at different zoom level positions pointing at the 3D map 8510. In this stage 8503, the user has provided input to decrease the zoom level of the 3D map 8510 by touching the screen of the device with two fingers away from each other and moving the fingers closer together while the fingers are touching the screen (e.g., a pinch gesture).
[0566] The zoom out adjustment is achieved by moving the virtual camera 8512 farther away from the 3D map 8510 along the line 8555. In some embodiments, the line 8555 along which the mapping application moves the virtual camera 8512 is a line formed by the front of the virtual camera 8512 and the focal point of the virtual camera 8512. The mapping application of some embodiments moves the virtual camera 8512 along the line formed by the front of the virtual camera 8512 and a location in the 3D map 8510 based on user input to zoom in on the view of the 3D map 8510.
[0567] Rendering a 3D map view using virtual camera 8512 at this location results in 3D map view 8534 in which buildings and roads appear farther away than the location shown in 3D map view 8524. As shown by the dashed version of virtual camera 8512, virtual camera 8512 moves farther away from 3D map 8510 along line 8555.
[0568] As mentioned above, Figure 85 Several exemplary zoom adjustment operations and corresponding movement of a virtual camera in a 3D map are shown, which are used to render a 3D map view of a 3D map. Those skilled in the art will recognize that any number of different zoom adjustments are possible. Furthermore, the mapping applications of some embodiments perform zoom adjustment operations in response to additional and / or different types of input (e.g., tapping the screen, double-tapping the screen, etc.).
[0569] Figure 86 A process 8600 of some embodiments is conceptually illustrated for adjusting the position of a virtual camera used to render a map view in response to a zoom level adjustment. In some embodiments, when a mapping application is in a map viewing mode (e.g., a location browsing mode, a navigation mode, etc.) and the mapping application receives a gesture via a touch screen of a device on which the mapping application is executing, the above and below references Figure 1-108 The described mapping application performs process 8600.
[0570] Process 8600 begins by receiving (at 8610) a gesture for adjusting the zoom level of a map view. In some embodiments, the gesture for adjusting the zoom level of a map view comprises a multi-touch gesture (e.g., touching the touch screen with multiple fingers simultaneously) received via the touch screen, such as described above with reference to FIG. Figure 85 In this example, process 8600 receives a two-touch pinch or expand gesture.
[0571] Next, process 8600 determines (at 8620) a zoom location based on the received gesture. In some embodiments, the zoom location is a location in the map toward which the zoom level is adjusted (e.g., for zooming in) or away from which the zoom level is adjusted (e.g., for zooming out). In some embodiments, process 8600 determines the zoom location by (1) identifying the locations of the initial two touch points of the gesture, and (2) identifying the zoom location as a midpoint along a line formed by the locations of the two touch points. In some embodiments, process 8600 uses additional techniques and / or different techniques to determine the zoom location. For example, in some embodiments, process 8600 uses the center of the map view as the zoom location.
[0572] The process 8600 then determines (at 8630) a zoom amount based on the received gesture. The process 8600 of some embodiments determines the zoom amount based on the difference in length between: (1) a first line formed by the locations of the initial two touch points of the gesture (e.g., where the user touched the touch screen with two fingers) and (2) a second line formed by the locations of the two touch points at the end of the gesture (e.g., where the user lifted the two fingers off the touch screen). In some embodiments, the process 8600 determines a positive zoom amount (e.g., a zoom-in adjustment) when the length of the second line is greater than the length of the first line, and determines a negative zoom amount (e.g., a zoom-out adjustment) when the length of the second line is less than the length of the first line. When the difference between the length of the second line and the length of the first line is equal to or within a threshold length amount, the process 8600 of some embodiments determines the zoom amount to be zero (e.g., no zoom adjustment). In some embodiments, the process 8600 uses additional techniques and / or different techniques to determine the zoom amount.
[0573] Next, the process 8600 adjusts (at 8640) the zoom level of the map view based on the identified zoom position and zoom amount. Figure 3-5 and Figures 74-80 As described, in some embodiments, a virtual camera is used to render a map view. Process 8600 of some embodiments adjusts the zoom level of the map view by moving the position of the virtual camera (1) along a line formed by the front of the virtual camera and the zoom position and (2) by an amount based on the zoom amount.
[0574] Finally, process 8600 renders (at 8650) the map view for display based on the adjusted zoom level of the map view. Process 8600 then ends.
[0575] 6. Virtual Camera Movement During Rotation
[0576] Figure 87 The following conceptually illustrates the rotation operation of a 3D map view according to some embodiments of the present invention. Specifically, Figure 87 The rotation operation is shown in two different stages 8705 and 8710. In some embodiments, the mapping application is referenced by Figure 90 To perform the process 9000 described below to perform the rotation operation.
[0577] Stage 1 8705 shows a 3D map view 8715 of a diagonal intersection of two roads and three buildings near the intersection. As shown, the three buildings are located at the top and sides of the diagonal intersection in 3D map view 8715. Stage 1 8705 also shows a user performing a gesture to rotate 3D map view 8715. As shown, the user has used two fingers to touch two locations near the center of 3D map view 8715 on the touch screen and has rotated the two locations in a clockwise direction.
[0578] The second stage 8710 shows the corresponding movement of the virtual camera 8700 around the 3D map 8725 in response to the rotation gesture shown in the first stage 8705, and the corresponding movement is used to render the 3D map view 8715. In this example, the rotation gesture is received near or at the intersection of two roads. As shown, the mapping application has moved the virtual camera 8700 from the left side of the 3D map 8725 toward the bottom of the 3D map 8725 by rotating the virtual camera 8700 around the Z axis located at or near the center of the rotation gesture. The 3D map view 8720 shows an exemplary rendering of the 3D map 8725 after the rotation of the virtual camera 8700. As shown, the 3D map view is now rotated so that the three buildings and the road rotate in a clockwise direction in response to the clockwise rotation gesture shown in the first stage 8705.
[0579] Although Figure 87 The rotation gesture is shown as being performed near or at the center of the 3D map view, but the rotation gesture can be performed at any location in the displayed 3D map view. In some embodiments, the mapping application rotates the 3D map based entirely or in part on the location of the rotation gesture. For example, if a rotation gesture is performed in the upper right corner of the 3D map view, the mapping application of some embodiments rotates the virtual camera about a Z-axis located at or near a location in the 3D map corresponding to the location of the rotation gesture.
[0580] In some embodiments, the mapping application provides a location indicator feature that, when activated, repositions the 3D map view at the user's location (e.g., the device's location in the 3D map). When rotation of the 3D map view is performed subsequent to activation of the location indicator feature, the mapping application of some such embodiments performs rotation of the 3D map view in different manners.
[0581] Figure 88 The following conceptually illustrates the rotation operation of a 3D map view after activating such a location indicator feature according to some embodiments of the present invention. Specifically, Figure 88 The rotation operation is shown in two different stages 8805 and 8810.
[0582] The first stage 8805 shows a 3D map view 8815 of a diagonal intersection of two roads and three buildings near the intersection. As shown, the three buildings are located at the top and side of the diagonal intersection in the 3D map view 8815. The first stage 8805 shows the 3D map view after the user has activated the location indicator feature, which causes the mapping application to reposition the 3D map view on the user's location, which is represented by a circle in this example. In addition, the first stage 8805 shows that the user is performing a gesture for rotating the 3D map view 8815 in the upper portion of the 3D map view 8815. As shown, the user has touched two locations in the upper portion of the 3D map view 8815 on the touch screen using two fingers and has rotated the two locations in a clockwise direction.
[0583] The second stage 8810 shows corresponding movement of the virtual camera 8800 around the 3D map 8825 in response to the rotate gesture shown in the first stage 8805, which is used to render the 3D map view 8815. In this example, the rotate gesture is received in the upper portion of the 3D map view 8815 behind the middle building after the location indicator feature is activated.
[0584] In some embodiments, when the mapping application receives a rotate gesture after the location indicator feature is activated, the mapping application rotates the 3D map around the user's location. The mapping application of some embodiments performs a rotate gesture only if the mapping application receives a rotate gesture after the location indicator feature has been activated and the mapping application has not received any intervening non-rotation gesture. Figure 88 In some embodiments, the mapping application performs the following rotation operation after activating the location indicator feature: Figure 88 The rotation shown operates for a defined amount of time and then resumes Figure 87 The rotation operation shown.
[0585] As shown in the second stage 8810, the mapping application has moved the virtual camera 8800 from the left side of the 3D map 8825 toward the bottom of the 3D map 8825 by rotating the virtual camera 8800 about the Z axis located at or near the user's location, indicated by a black dot, at the intersection of two roads. 3D map view 8820 shows an exemplary rendering of the 3D map 8825 after rotating the virtual camera 8800. As shown, the 3D map view is now rotated so that the three buildings and the road rotate in a clockwise direction around the user's location in response to the clockwise rotation gesture shown in the first stage 8805.
[0586] above Figure 87 and Figure 88 Shown is rotating the 3D map while viewing the 3D map in the 3D perspective view.In some embodiments, the mapping application also allows the user to rotate the 3D map while viewing the 3D map in the 2D perspective view. Figure 89 The following conceptually illustrates a rotation operation of a 3D map view from a 2D perspective view of the 3D map according to some embodiments of the present invention. Specifically, Figure 89 The rotation operation is shown in three different stages 8905-8915 along with the inertial effects of the rotation operation. In some embodiments, the mapping application is referenced by Figure 90 To perform the process 9000 described below to perform the rotation operation.
[0587] Stage 1 8905 shows a 3D map view 8920 of a 3D map. As shown, 3D map view 8920 includes several streets extending in parallel or perpendicular directions. Stage 1 8905 also shows that a user is providing input for rotating 3D map view 8920. Specifically, the user is performing a gesture for rotating 3D map view 8920 by touching two fingers to two locations on the touch screen and rotating the two fingers in a clockwise direction as indicated by the two arrows.
[0588] The second stage 8910 shows the 3D map immediately after the user has completed input for rotating the 3D map. For this example, the user has completed the input by lifting two fingers off the touch screen of the device. In addition, the second stage 8910 shows a 3D map view 8925 of the 3D map rendered by the mapping application. As shown, the mapping application has rotated the 3D map in a c...
Claims
1. A mobile device, comprising: A set of processing units for executing instruction sets; A non-transitory machine-readable medium storing a mapping application that, when executed on the mobile device by at least one of the processing units, provides an animated navigation presentation showing the device navigating along a 3D map of an area the device is traversing, the mapping application comprising a set of instructions for: defining a virtual camera that identifies a view of the 3D map to render at each of a plurality of time instances; As the device traverses the area, generating a navigation presentation by: moving the virtual camera along a navigated route in the 3D map to identify different views of the 3D map for rendering; As the navigation presentation approaches a first threshold distance along the navigated route to a turn, moving the virtual camera upward from a 3D perspective view to a top-down two-dimensional view of the map at the location of the turn; As the navigation presentation passes a second threshold distance after the turn along the navigated route, the virtual camera is moved downward from the top-down view to a 3D perspective view of the map.
2. A mobile device according to claim 1, wherein the mapping application also includes an instruction set for: displaying a representation of the device traveling along the navigated route as the device travels through the area, wherein when the displayed representation of the device reaches the first threshold distance of the turn, the navigation presents the first threshold distance to the turn, and when the displayed representation of the device passes the second threshold distance, the navigation presents the second threshold distance through the turn.
3. A mobile device according to claim 1, wherein the instruction set for moving the virtual camera downward includes an instruction set for: moving the camera downward so that the virtual camera can define the following view of the 3D map, the view including more locations on the map in front of the direction of travel of the device. 4 . The mobile device of claim 1 , wherein a focus of the virtual camera is a defined distance in front of a location of a representation of the device on the 3D map. The mobile device of claim 4 , wherein the distance from the focal point depends on the speed at which the device is traveling.
6. The mobile device of claim 4, wherein the mapping application further comprises a set of instructions for: While moving the virtual camera upward, rotating the virtual camera toward the 3D map; and While moving the virtual camera downward, the virtual camera is rotated away from the 3D map.
7. The mobile device of claim 4, wherein the virtual camera tilts downward to maintain the focus while moving along the navigated route.
8. A non-transitory machine-readable medium storing a mapping application that, when executed by at least one processing unit of a device, provides an animated navigation presentation showing the device navigating along a 3D map of an area the device is traversing, the mapping application comprising a set of instructions for: defining a virtual camera that identifies a view of the 3D map to render at each of a plurality of time instances; As the device traverses the area, generating a navigation presentation by: moving the virtual camera along a navigated route in the 3D map to identify different views of the 3D map for rendering; As the navigation presentation approaches a first threshold distance along the navigated route to a turn, moving the virtual camera upward from a 3D perspective view to a top-down two-dimensional view of the map at the location of the turn; As the navigation presentation passes a second threshold distance after the turn along the navigated route, the virtual camera is moved downward from the top-down view to a 3D perspective view of the map.
9. According to the non-temporary machine-readable medium according to claim 8, the mapping application also includes an instruction set for: as the device traverses the area, displaying a representation of the device traversing along the navigated route, wherein when the displayed representation of the device reaches the first threshold distance of the turn, the navigation presents the first threshold distance to the turn, and when the displayed representation of the device passes the second threshold distance, the navigation presents the second threshold distance through the turn.
10. The non-transitory machine-readable medium of claim 8, wherein the mapping application further comprises the following set of instructions: While moving the virtual camera upward, rotating the virtual camera toward the 3D map; and While moving the virtual camera downward, the virtual camera is rotated away from the 3D map.
11. A non-transitory machine-readable medium according to claim 8, wherein the instruction set for moving the virtual camera downward includes an instruction set for: moving the virtual camera downward so that the virtual camera can define a view of the 3D map as follows, the view including more locations on the map in front of the direction of travel of the device. 12 . The non-transitory machine-readable medium of claim 8 , wherein a focus of the virtual camera is a defined distance in front of a location of the representation of the device on the 3D map.
13. The non-transitory machine-readable medium of claim 12, wherein the distance from the focal point depends on a speed at which the device is traveling.
14. The non-transitory machine-readable medium of claim 12, wherein the virtual camera tilts downward to maintain the focus while moving along the navigated route.
15. A method of providing an animated navigation presentation along a 3D map of an area a device is traversing, the method comprising: defining a virtual camera that identifies a view of the 3D map to render at each of a plurality of time instances; As the device traverses the area, generating a navigation presentation by: moving the virtual camera along a navigated route in the 3D map to identify different views of the 3D map for rendering; As the navigation presentation approaches a first threshold distance along the navigated route to a turn, moving the virtual camera upward from a 3D perspective view to a top-down two-dimensional view of the map at the location of the turn; As the navigation presentation passes a second threshold distance after the turn along the navigated route, the virtual camera is moved downward from the top-down view to a 3D perspective view of the map.
16. The method according to claim 15, further comprising: As the device traverses the area, a representation of the device traversing along the navigated route is displayed, wherein when the displayed representation of the device reaches the first threshold distance of the turn, the navigation presents the first threshold distance to the turn, and when the displayed representation of the device passes the second threshold distance, the navigation presents the second threshold distance through the turn.
17. The method according to claim 15, further comprising: While moving the virtual camera upward, rotating the virtual camera toward the 3D map; as well as While moving the virtual camera downward, the virtual camera is rotated away from the 3D map.
18. The method of claim 15, wherein moving the virtual camera downward comprises: The virtual camera is moved downward so that the virtual camera can define a view of the 3D map that includes more locations on the map that are forward in the direction of travel of the device.
19. The method of claim 15, wherein a focus of the virtual camera is a defined distance in front of a location of the representation of the device on the 3D map.
20. The method of claim 19, wherein the distance from the focal point depends on the speed at which the device is traveling.
Citation Information
Patent Citations
Context-aware voice guidance
US20130322634A1
Problem Reporting in Maps
US20130326407A1
Scalable and Efficient Cutting of Map Tiles
US20130328879A1
Scalable processing for associating geometries with map tiles
US9235906B2
Virtual camera for 3d maps
CN107155342A
