Method and device for constructing local map in driving process of mobile equipment

Abstract

The embodiment of the invention discloses a method and device for constructing a local map in the driving process of mobile equipment, and the method comprises the steps: determining a first perception result based on an environment image of the mobile equipment in the driving process; according to the first sensing result, determining physical object information of the mobile device under a local coordinate system; determining logic element information of a local map to be constructed according to the physical object information; the logic element information is used for representing a road logic association relationship; and according to the logic element information, constructing a local map of the preset range of the current position of the mobile device. According to the embodiment of the invention, on the basis of reconstructing the surrounding physical elements, the logic element information capable of representing the road logic association relationship can be determined and used for constructing the local map, so that the semantic information richness of the local map is effectively improved, and the requirements of navigation and path planning can be better met.

Description

Technical Field

[0001] The present disclosure relates to computer vision technology, and more particularly to a method and device for constructing a local map during the driving process of a movable device. Background Art

[0002] In intelligent driving scenarios, vehicles usually need to navigate and plan routes through local maps within a certain range around the vehicle while driving. In related technologies, vehicles reconstruct surrounding physical elements (or physical objects) through the perception of the surrounding environment during driving for downstream navigation and route planning. However, the semantic information richness of physical elements is low and often cannot meet the needs of navigation and route planning. Summary of the invention

[0003] The embodiments of the present disclosure provide a method and device for constructing a local map during the driving process of a movable device. On the basis of reconstructing the surrounding physical elements, logical element information that can characterize the logical association relationship of roads can be determined for constructing the local map, thereby effectively improving the semantic information richness of the local map to better meet the needs of navigation and path planning.

[0004] A first aspect of an embodiment of the present disclosure provides a method for constructing a local map during the driving process of a movable device, comprising: determining a first perception result based on an environmental image of the movable device during the driving process; determining physical object information in a local coordinate system where the movable device is located according to the first perception result; determining logical element information of a local map to be constructed according to the physical object information; the logical element information is used to characterize a logical association relationship of roads; and constructing a local map of a preset range of a current position of the movable device according to the logical element information.

[0005] A second aspect of an embodiment of the present disclosure provides a device for constructing a local map during the driving of a movable device, including: a first processing module, used to determine a first perception result based on an environmental image of the movable device during the driving process; a second processing module, used to determine physical object information in a local coordinate system where the movable device is located according to the first perception result; a third processing module, used to determine logical element information of the local map to be constructed according to the physical object information; the logical element information is used to characterize the logical association relationship of roads; and a fourth processing module, used to construct a local map of a preset range of the current position of the movable device according to the logical element information.

[0006] According to a third aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the storage medium stores a computer program for executing the method for constructing a local map during driving of a movable device as described in any of the above embodiments of the present disclosure.

[0007] According to a fourth aspect of an embodiment of the present disclosure, an electronic device is provided, comprising: a processor; a memory for storing executable instructions of the processor; and the processor for reading the executable instructions from the memory and executing the instructions to implement the method for constructing a local map during driving of a movable device as described in any of the above embodiments of the present disclosure.

[0008] A fifth aspect of the embodiments of the present disclosure provides a computer program product. When the instructions in the computer program product are executed by a processor, the method for constructing a local map during the driving of a movable device provided in any of the above embodiments of the present disclosure is performed.

[0009] Based on the method and device for constructing a local map during the driving of a movable device provided by the above-mentioned embodiments of the present disclosure, the logical element information of the local map to be constructed can be determined on the basis of reconstructing the physical elements around the movable device. Since the logical element information can characterize the logical association relationship of roads and has richer semantic information, the logical element information is used to construct the local map. Therefore, the richness of the semantic information of the local map can be effectively improved, and a more effective local map can be provided for the navigation and path planning of the movable device. On the basis of avoiding the use of pre-built high-precision maps or crowdsourced maps, the autonomous driving of the movable device can be realized based on a pure perception method. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 This is an exemplary application scenario of the method for constructing a local map during the driving of a mobile device provided by the present disclosure;

[0011] Figure 2 It is a flowchart of a method for constructing a local map during driving of a movable device provided by an exemplary embodiment of the present disclosure;

[0012] Figure 3 is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0013] Figure 4 is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0014] Figure 5 is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0015] Figure 6 is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0016] Figure 7 is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0017] Figure 8 is a flowchart of a method for constructing a local map during driving of a mobile device provided by an exemplary embodiment of the present disclosure;

[0018] Fig. 9 is an exemplary schematic diagram of a lane line rough grouping result provided by an exemplary embodiment of the present disclosure;

[0019] Fig.10 is another exemplary schematic diagram of a lane line rough grouping result provided by an exemplary embodiment of the present disclosure;

[0020] Fig.11 is a lane line subdivision schematic diagram provided by an exemplary embodiment of the present disclosure;

[0021] Fig.12 is a lane line clipping schematic diagram provided by an exemplary embodiment of the present disclosure;

[0022] Fig.13 is a visual schematic diagram of a logic element provided by an exemplary embodiment of the present disclosure;

[0023] Fig.14 is a visual schematic diagram of a logic element provided by another exemplary embodiment of the present disclosure;

[0024] Fig.15 is a schematic structural diagram of an apparatus for constructing a local map during the driving of a movable device provided by an exemplary embodiment of the present disclosure;

[0025] Fig.16 is a structural schematic diagram of an apparatus for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0026] Fig.17 is a structural schematic diagram of an apparatus for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure;

[0027] Fig.18 It is a structural diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0028] To explain the present disclosure, example embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. It should be understood that the present disclosure is not limited to the example embodiments.

[0029] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0030] SUMMARY OF THE DISCLOSURE

[0031] In the process of realizing the present disclosure, the inventors found that in intelligent driving scenarios, vehicles usually need to navigate and plan routes through local maps within a certain range around the vehicle during driving. In related technologies, during driving, the vehicle reconstructs the surrounding physical elements through the perception results of the surrounding environment, such as lane lines, curbs, stop lines, crosswalks and other real physical elements in the environment, for downstream navigation and route planning. However, the semantic information richness of physical elements is low and often cannot meet the needs of navigation and route planning.

[0032] Exemplary Overview

[0033] Figure 1 This is an exemplary application scenario of the method for constructing a local map during the driving of a mobile device provided by the present disclosure. Figure 1As shown, the movable device takes a vehicle (or self-vehicle) 11 as an example. In actual applications, the movable device can also be a robot or other movable device. When the vehicle 11 is driving on the road, one or more cameras 12 installed on the vehicle 11 can collect the surrounding environment image for the construction of the local map. Specifically, the first perception result can be determined based on the environmental image of the vehicle 11 during driving; according to the first perception result, the physical object information in the local coordinate system (such as the xoy coordinate system in the figure) where the vehicle 11 is located is determined. The physical object information may include, for example, the object information of one or more physical objects such as lane lines 13, curbs 14, stop lines 15, zebra crossings 16, etc. in the local coordinate system. The object information may include, for example, one or more of the position, direction, size and other information of the physical object. The specific physical object information is not limited. Then, the logical element information of the local map to be constructed can be determined based on the obtained physical object information. Logical element information is used to characterize the logical association of roads, and can also be used to characterize semantic information of roads. For example, logical element information may include element information of one or more elements such as lane elements, driving line elements, intersection elements, virtual lane elements in intersections, etc. The element information may include, for example, one or more of the element's identification, position, direction, size, semantics, physical object information involved in or included in the element, and the association relationship between elements. Then, a local map of a preset range around the current position of the vehicle 11 can be constructed based on the logical element information. Since the logical element information includes richer semantic information, the semantic richness of the constructed local map can be effectively improved, thereby providing a more effective local map for navigation and path planning of the mobile device, and autonomous driving of the mobile device can be achieved based on pure perception while avoiding the use of pre-constructed high-precision maps or crowdsourced maps.

[0034] In practical applications, the first perception result can be obtained by combining the environment image and other types of sensor data. Other types of sensor data may include, for example, radar, lidar, millimeter wave radar, ultrasonic radar, and other types of sensor data.

[0035] Exemplary Methods

[0036] Figure 2 1 is a flow chart of a method for constructing a local map during the driving of a mobile device provided by an exemplary embodiment of the present disclosure. This embodiment can be applied to an electronic device, such as a computing platform (or mobile device terminal) on a mobile device, such as Figure 2 As shown, the method of the embodiment of the present disclosure may include the following steps:

[0037] Step 210: determining a first perception result based on an image of the environment of the movable device during driving.

[0038] Among them, the movable device can be an intelligent driving device such as a vehicle or a robot that needs to navigate and plan a path based on a local map. The environmental image can be obtained based on the camera set on the movable device. The first perception result can be obtained by perceiving and processing the environmental image based on a preset perception algorithm and / or a preset perception model. The first perception result may include the perception result of at least one physical object. A physical object refers to an object that actually exists in the environment around the movable device. For example, a physical object may include continuous (or linear) objects such as lane lines and curbs, as well as one or more discrete (i.e., can be represented by a point or an area) objects such as stop lines, crosswalks, traffic lights, cones, arrows, etc. The specific perceptible physical objects are not limited. The perception result of the physical object may include the target detection result, semantic segmentation result, etc. of the physical object in the image coordinate system of the environmental image. Optionally, the perception result of the physical object may also include the perception result in the local coordinate system of the movable device obtained by further processing the target detection result, semantic segmentation result, etc.

[0039] In some optional embodiments, the first perception result can be obtained based on a perception function system set on the mobile device, which is not specifically limited.

[0040] In some optional embodiments, the environment image may include an image captured by at least one camera.

[0041] Step 220: Determine the physical object information in the local coordinate system where the movable device is located based on the first perception result.

[0042] Among them, the local coordinate system where the movable device is located is a coordinate system with a preset position on the movable device as the origin. Taking the movable device as a vehicle as an example, the local coordinate system can be a vehicle coordinate system (Vehicle Coordinate System, abbreviated as: VCS, or vehicle coordinate system) with the center of the rear axle of the vehicle as the origin. In practical applications, the preset position is not limited to the center of the rear axle. Physical object information refers to relevant information about the perceived physical object. For example, physical object information can include an ordered set of points of each perceived physical object (such as lane lines, curbs, stop lines, crosswalks, etc.) in the local coordinate system.

[0043] In some optional embodiments, when the first perception result includes target detection results and semantic segmentation results of the physical object in the image coordinate system, the first perception result can be converted and processed accordingly based on the conversion relationship between the image coordinate system and the local coordinate system to obtain the physical object information in the local coordinate system. When the first perception result includes the physical object information in the local coordinate system, the physical object information can be directly obtained from the first perception result. The specific method of obtaining the physical object information is not limited.

[0044] Step 230, determining the logical element information of the local map to be constructed based on the physical object information; the logical element information is used to characterize the logical association relationship of the roads.

[0045] Among them, the logical elements may include elements with different semantic types from the physical objects abstracted based on the physical objects. The logical element information may include one or more of the ordered point set, position, size, direction, semantics, and other information of the logical elements in the local coordinate system. Exemplarily, the logical element information may include lane element information, driveline element information, junction element information, virtual lane element information in the junction, virtual driveline element information, and the like. The lane element is a logical element abstracted from the part between two adjacent lane line objects. The driveline element is a logical element abstracted from the center line between two adjacent lane line objects. The virtual lane element is a logical element abstracted from the junction based on the association between lanes around the junction. For example, the virtual lane element may include a straight lane element, a left turn lane element, a right turn lane element, and the like. Since there is actually no lane in the junction, it is called a virtual lane element. The virtual driveline element is a logical element abstracted from the center line of the virtual lane in the junction. Intersection elements are logical elements abstracted from the intersection area and the logical elements (such as virtual lanes, virtual driving lines) and physical objects (such as stop lines, crosswalks) contained in the intersection. Logical element information can be used to represent road logical associations and semantic information.

[0046] In some optional embodiments, logical element information corresponding to the physical object may also be determined based on the physical object information to construct a local map.

[0047] Step 240, constructing a local map of a preset range of the current position of the movable device according to the logical element information.

[0048] The preset range may be set according to the range that can be perceived by the environment image. For example, the preset range may be an area range within a preset distance around the origin of the local coordinate system. For example, the preset range may be an area range within 200 meters around and 50 meters around the origin of the local coordinate system. The specific range is not limited.

[0049] In some optional embodiments, a local map within a preset range around the current position of the movable device may be constructed based on the association relationship between the logical elements in the logical element information.

[0050] In some optional embodiments, a local map may be constructed based on physical object information and logical element information to ensure semantic richness of the local map.

[0051] In some optional embodiments, each physical object can be used as the information content of the logical element associated with the physical object, that is, the logical element information of each logical element can include the physical object information constituting the logical element. For example, the lane element information can include the lane line object information constituting the lane element. The intersection element information can include the stop line, crosswalk and other physical object information in the intersection. In this way, the physical objects and logical objects of the constructed local map can have an accurate mapping relationship.

[0052] The method for constructing a local map during the driving of a movable device provided in this embodiment can determine the logical element information of the local map to be constructed based on the reconstruction of the physical elements around the movable device. Since the logical element information can characterize the logical association relationship of roads and has richer semantic information, the physical object information can be combined with the logical element information to construct a local map that is more suitable for use by the movable device. The local map can have rich semantic information, providing a more effective local map for the navigation and path planning of the movable device. The autonomous driving of the movable device can be achieved based on a pure perception method without using pre-constructed maps (such as high-precision maps, crowdsourced maps, etc.). In the absence of a map, the driving functions such as going straight, changing lanes, turning left or right at intersections, U-turns, and going straight can be achieved autonomously.

[0053] In some optional embodiments, step 210, determining the first perception result based on the environment image of the movable device during driving, may include: determining the first perception result based on the environment image and at least one other sensor data. The other sensor data may include radar sensor data, laser radar sensor data, millimeter wave radar sensor data, ultrasonic radar sensor data, etc.

[0054] Figure 3 It is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure.

[0055] In some optional embodiments, in the above Figure 2 Based on the embodiment shown, Figure 3As shown, step 230 of determining the logical element information of the local map to be constructed according to the physical object information may include:

[0056] Step 2310: Determine lane line object information in the environment where the movable device is located based on the physical object information.

[0057] The lane line object information may include an ordered point set for each lane line in at least one lane line. The order of points in the ordered point set of the lane line may be an order according to the drivable direction of the road where the lane line is located. For example, if the lane is a lane from east to west, the ordered point set of the lane line on the lane is a sequence point set from east to west. The lane line object information may be extracted from the physical object information based on the shape points and geometric information of the physical object.

[0058] Step 2320: Determine the lane element information and driving line element information required for the local map based on the lane line object information.

[0059] Among them, the lane element is a logical element abstracted from the part between two adjacent lane line objects. The driving line element is a logical element abstracted from the center line between two adjacent lane line objects. The lane element information may include one or more attribute information of the lane element and the logical association relationship between the lane element and other elements. The attribute information of the lane element may include, for example, the lane element identifier (or may be called the lane element unique identifier), the lane line object identifier constituting the lane element, other object identifiers within the area of ​​the lane element (such as arrow objects), and so on. The logical association relationship between the lane element and other elements may include, for example, the association relationship along the extension direction of the lane, the association relationship perpendicular to the extension direction of the lane (i.e., the left and right adjacency relationship of the lane), the parent-child relationship, and so on. The driving line element information may include one or more attribute information of the driving line element. The attribute information of the driving line element may include, for example, the driving line element identifier, the ordered point set constituting the driving line element, the direction of the driving line element (which may be expressed as the direction angle in the local coordinate system), and other information. The logical element information corresponding to different logical elements may be different.

[0060] Step 2330, determining the logical element information of the local map based on the lane element information and the driving line element information.

[0061] In some optional embodiments, for example, in non-intersection scenarios, lane element information and driving line element information can be determined as logical element information of a local map. Alternatively, more element information of other semantic types can be further abstracted based on the lane element information to further expand the semantic richness of the logical element information. For example, lane segment element information can be abstracted based on the lane element information. Specifically, multiple adjacent lane elements with the same drivable direction can be formed into lane segment elements to obtain lane segment element information.

[0062] In some optional embodiments, such as in intersection scenarios, it is also possible to determine intersection-related element information based on the lane element information and driving line element information, such as virtual lane element information, virtual lane segment element information, virtual driving line element information, etc. within the intersection, and then determine the logical element information of the local map by integrating element information of various semantic types.

[0063] The disclosed embodiments can add lane semantic information and driving line semantic information to a local map by abstracting lane elements and driving line elements from lane line objects, thereby improving the semantic richness of the local map. The lane elements represent the lane areas where the movable device can drive, and the driving line elements represent the reference lines where the movable device can drive on the lanes. Therefore, effective map reference information can be provided for navigation and path planning of the movable device.

[0064] Figure 4 It is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure.

[0065] In some optional embodiments, such as Figure 4 As shown, step 2320 of determining lane element information and driving line element information required for the local map according to the lane line object information may include:

[0066] Step 23210, determining the preset direction type of each lane line based on the lane line object information.

[0067] Among them, the preset direction type may include a transverse type, a longitudinal type, an oblique type, etc. The transverse type refers to a type that is perpendicular to the current driving direction of the mobile device (or the current lane), and the longitudinal type refers to a type along the current driving direction of the mobile device (or the current lane). The oblique type may further include a type that is different from the lane where the mobile device is currently located (which may be referred to as a first oblique type and a second oblique type). For example, a type that is oblique from the right front to the left rear of the mobile device (a first oblique type), and a type that is oblique from the left front to the right rear of the mobile device (a second oblique type). The specific preset direction type may be set according to the intersection type. If the intersection is a crossroads, the preset direction type includes a transverse type and a longitudinal type. If the intersection type is a Y-shaped intersection, the preset direction type includes a longitudinal type and an oblique type. Each lane line may correspond to a preset direction type. The preset direction type of the lane line may be determined according to the direction angle of the lane line in the local coordinate system or the angle between the lane line and the longitudinal axis (or transverse axis) of the local coordinate system. For example, the preset direction type of the lane line can be determined as the horizontal type according to the angle between the lane line and the longitudinal axis (X axis) of the local coordinate system being close to 90 degrees, and the preset direction type of the lane line can be determined as the longitudinal type according to the angle between the lane line and the longitudinal axis (X axis) of the local coordinate system being close to 0 degrees. The lane line can be determined as the oblique type according to the angle between the lane line and the longitudinal axis (X axis) of the local coordinate system being close to 45 degrees or 135 degrees.

[0068] In some optional embodiments, each lane line may be clustered according to its direction angle in the lane line object information to obtain clusters of different direction types, and a preset direction type of the lane lines in the cluster may be obtained according to the direction type of each cluster.

[0069] Step 23220: based on the preset direction type of each lane line, group the lane lines in the lane line object information to obtain at least one lane line group.

[0070] Among them, lane lines of the same preset direction type are determined as a lane line group. For example, if the preset direction type includes a horizontal type and a vertical type, all lane lines belonging to the horizontal type are determined as a lane line group (which may be called a horizontal lane line group), and all lane lines belonging to the vertical type are determined as a lane line group (which may be called a vertical lane line group).

[0071] In some optional embodiments, before grouping, each lane line in the lane line object information may be filtered according to the lane line length, and lane lines with a length less than a preset length threshold are filtered out to reduce the adverse effects of noise.

[0072] In some optional embodiments, each lane line can be clustered according to its direction angle in the lane line object information to obtain clusters of different preset direction types, and the cluster can be determined as a lane line group of the corresponding preset direction type according to the preset direction type of each cluster.

[0073] Step 23230: For each lane line group in at least one lane line group, cluster the lane line group to obtain at least one cluster corresponding to the lane line group.

[0074] For each lane line group, the lane line group may be clustered based on the center points of each lane line in the lane line group to obtain at least one cluster corresponding to the lane line group. The center point of the lane line may be determined based on an ordered point set of the lane line. For example, a point in the middle of the ordered point set of the lane line is used as the center point of the lane line.

[0075] In some optional embodiments, each lane line group may be clustered based on density clustering or other clustering methods to obtain at least one cluster corresponding to the lane line group. The purpose of clustering is to cluster lane lines with similar center points into the same cluster. For example, in an intersection scenario, taking a crossroad as an example, the horizontal type lane line group around the intersection is clustered into at least two horizontal clusters due to the separation of the intersection. The vertical type lane line group, which is separated by the intersection, can be clustered into at least two vertical clusters.

[0076] Step 23240, based on the cluster corresponding to each lane line group, determine the lane element information and the driving line element information.

[0077] Among them, for each cluster corresponding to each lane line group, the lane line feature information and driving line feature information corresponding to the cluster can be determined. Based on the lane line feature information and driving line feature information corresponding to each cluster of each lane line group, the lane line feature information and driving line feature information required for the local map are determined. The lane line feature information and driving line feature information corresponding to each cluster are determined based on the lane line (i.e., lane line object) information in the cluster. The method of determining the lane line feature information and driving line feature information based on the lane line object information refers to the aforementioned embodiment and will not be repeated here.

[0078] In the disclosed embodiment, lane lines are divided into different lane line groups according to their preset direction types, and the lane line groups are clustered. This ensures the continuity of each lane line in each cluster and avoids the influence of areas such as intersections that affect lane elements and driving line elements, thereby improving the accuracy and effectiveness of lane element information and driving line information.

[0079] In some optional embodiments, step 2320 of determining lane element information and driving line element information required for the local map based on the lane line object information may include: determining the direction angle of each lane line based on the lane line object information; clustering each lane line based on the direction angle of each lane line to obtain at least one cluster; determining a preset direction type for each cluster; treating each cluster as a lane line group of a preset direction type; clustering each lane line group to obtain at least one cluster corresponding to the lane line group; and determining the lane element information and driving line element information based on the cluster corresponding to each lane line group.

[0080] Among them, the direction angle of each lane line can be the direction angle of the lane line in the local coordinate system or the angle between the lane line and the longitudinal axis (or transverse axis) of the local coordinate system. Clustering is performed based on the direction angle of each lane line. Lane lines with the same or similar direction angles can be clustered into the same cluster, and lane lines with large direction angle differences can be divided into different clusters. Based on this, lane lines of different preset direction types (such as horizontal type, longitudinal type) can be divided into different lane line groups, such as horizontal lane line groups, longitudinal lane line groups, etc. The clustering method can be a density-based clustering method or other clustering methods, which are not specifically limited. After obtaining at least one lane line group, the specific operation of determining the lane element information and the driving line element information according to each lane line group can refer to the above steps 23230 and 23240.

[0081] In some optional embodiments, step 23230 of clustering the lane line group for each lane line group in at least one lane line group to obtain at least one cluster corresponding to the lane line group may include: for each lane line group in at least one lane line group, determining the center point of each lane line in the lane line group; clustering the lane line group based on the center point of each lane line to obtain at least one cluster corresponding to the lane line group.

[0082] The center point of each lane line can be determined based on the ordered point set of the lane line. For example, a point in the middle of the ordered point set of the lane line is used as the center point of the lane line, or the center point of the lane line is determined based on the position mean of the ordered point set. The specific method for determining the center point of the lane line is not limited.

[0083] In some optional embodiments, the clustering methods for clustering lane line groups may include DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering method, K-MEANS (K-means) clustering method, hierarchical clustering method, etc.

[0084] In some optional embodiments, for the lateral lane line group, clustering can be performed according to the horizontal coordinate of the center point of the lane line (the Y-axis coordinate of the local coordinate system). For the longitudinal lane line group, clustering can be performed according to the vertical coordinate of the center point of the lane line (the X-axis coordinate of the local coordinate system).

[0085] In the disclosed embodiment, since the center point of a lane line can represent the overall position of a continuous lane line, by clustering the center points of the lane lines, relatively clustered continuous lane lines can be formed into a cluster. For example, the lane lines on the east and west sides of an intersection are clustered into two clusters, thereby ensuring the consistency or similarity of the attributes of each lane line in the cluster, providing accurate and effective reference information for determining logical elements, thereby improving the accuracy and effectiveness of logical elements.

[0086] In some optional embodiments, determining lane element information and driving line element information based on the cluster corresponding to each lane line group in step 23240 may include:

[0087] Based on the clusters corresponding to each lane line group, the current scene type where the movable device is located is determined; based on the current scene type and the clusters corresponding to each lane line group, lane element information and driving line element information are determined.

[0088] The scene type may include intersection scenes, non-intersection scenes, and the like. The current scene type may be an intersection scene or a non-intersection scene. An intersection scene refers to a scene in which an intersection exists within a preset range of the current position of the mobile device, and a non-intersection scene refers to a scene in which no intersection exists.

[0089] In some optional embodiments, the current scene type where the movable device is located can be determined based on the clusters corresponding to each lane line group and the lane line characteristics around the intersection. For example, based on the preset direction type and cluster number of each lane line group, and the lane line groups and cluster numbers that should exist at different types of intersections, it is determined whether there is an intersection. If there are two lane line groups, a transverse lane line group and a longitudinal lane line group, and the transverse lane line group includes at least two clusters and / or the longitudinal lane line group includes at least two clusters, it is determined that there is an intersection. Taking an intersection as an example, the transverse lane line group includes at least two clusters, and the longitudinal lane line group includes at least two clusters. Taking a T-junction as an example, the transverse lane line group includes at least two clusters, and the longitudinal lane line group includes at least one cluster. Taking a Y-junction as an example, the longitudinal lane line group includes at least one cluster, and also includes at least two diagonal lane line groups, or includes at least two clusters of a diagonal lane line group.

[0090] In some optional embodiments, different scene types may have different physical objects. For example, in non-intersection scenes, lane lines will not be affected by intersections, nor will they affect intersections. However, in intersection scenes, the perceived lane lines may extend into the intersection. Therefore, different lane line processing methods may be set for different scene types. For example, for intersection scenes, the portion of the lane lines in the cluster that extend into the intersection needs to be cut off to ensure the accuracy of the lane lines, and then the lane element information and the driving line element information are determined based on the processed lane lines, so that the lane element information and the driving line element information can be more in line with the actual situation of the current scene type.

[0091] In the embodiments of the present disclosure, since different scene types involve different physical objects, the clustering of lane line groups needs to be processed differently. Therefore, in combination with the current scene type, the lane element information and the driving line element information can be determined in a manner that conforms to the current scene type, thereby ensuring the accuracy and effectiveness of the lane element information and the driving line element information.

[0092] In some optional embodiments, determining lane element information and driving line element information based on the current scene type and the cluster corresponding to each lane line group may include:

[0093] If the current scene type is an intersection scene, the stop line object information and intersection boundary information within the intersection are determined based on the physical object information; the lane element information and driving line element information are determined based on the clustering, stop line object information and intersection boundary information corresponding to each lane line group.

[0094] Wherein, in the intersection scene, the physical object information also includes the stop line object information and the intersection boundary information in the intersection, and the intersection boundary information is, for example, the intersection bounding box information obtained by perception. The intersection boundary information may include the center point and at least two corner points of the intersection bounding box, or the intersection boundary information may include the four corner points of the intersection bounding box, or the intersection boundary information may include the center point, direction angle and size of the intersection bounding box, or the intersection boundary information may include an ordered point set of the intersection bounding box. As long as the intersection boundary can be determined, the content of the specific intersection boundary information is not limited. The stop line object information may include information such as the center point, direction angle, length, etc. of the stop line, or the stop line object information may include an ordered point set of the stop line.

[0095] In some optional embodiments, the clusters can be further processed based on the positional relationship between the clusters corresponding to the lane line groups and the stop line objects and intersection boundaries. For example, the further processing is to cut off the portion of the lane lines in each cluster that extend into the intersection, and the processed clusters are used to determine the lane element information and the driving line element information.

[0096] The disclosed embodiment combines the cluster corresponding to each lane line group, the stop line object information in the intersection and the intersection boundary information to determine the lane element information and the driving line element information, so that the lane element information and the driving line element information can be more consistent with the actual element conditions in the intersection scene, further improving the accuracy and effectiveness of the logical element information.

[0097] In some optional embodiments, determining lane element information and driving line element information based on the cluster corresponding to each lane line group in step 23240 may include:

[0098] According to the direction angle of the lane lines in each cluster corresponding to each lane line group, the first lane line group in the first direction and the second lane line group in the second direction corresponding to each cluster are determined; according to the endpoints and virtual-real change points of the lane lines in the first lane line group corresponding to each cluster, the first lane line group is divided into at least one first sub-lane line group; according to the endpoints and virtual-real change points of the lane lines in the second lane line group corresponding to each cluster, the second lane line group is divided into at least one second sub-lane line group; based on the first sub-lane line group and the second sub-lane line group, the lane element information and the driving line element information are determined.

[0099] Among them, the direction angle of the lane line is the direction angle of the lane line in the local coordinate system, which can be expressed as the angle between the lane line direction and the longitudinal axis (X axis) of the local coordinate system. The lane line direction is the drivable direction of the road where the lane line is located. For example, if the drivable direction of the lane is from east to west, then the lane line direction is the direction from east to west. The direction angle of the lane line can be obtained from the physical object information. The first direction and the second direction represent different drivable directions on a two-way road. For example, if the lane where the mobile device is currently located is a lane that travels from south to north, for the transverse lane line in the local coordinate system, the first direction is the direction from east to west (relative to the direction from right to left of the mobile device), and the second direction is the direction from west to east (relative to the direction from left to right of the mobile device). For the longitudinal lane line in the local coordinate system, the first direction is the direction from south to north (that is, the driving direction of the mobile device, or the forward direction), and the second direction is the direction from north to south (that is, the reverse direction of the driving direction of the mobile device, or the reverse direction). According to the direction angle of the lane line, the lane lines of different drivable directions in each cluster are divided into different lane line groups to obtain a first lane line group in the first direction and a second lane line group in the second direction. For example, each cluster in the longitudinal lane line group can be divided into a first lane line group in the forward direction (first direction) relative to the movable device and a second lane line group in the reverse direction (second direction).

[0100] In some optional embodiments, the endpoints of each lane line include the starting point and the end point of the lane line. The endpoints of the lane line can be determined according to the ordered point set of the lane line, the first point in the ordered point set of each lane line is the starting point of the lane line, and the last point is the end point of the lane line. The virtual-real change point of the lane line includes at least one of the change point from the solid line to the dotted line and the change point from the dotted line to the solid line in the lane line. For example, a lane line includes a first solid line segment, a dotted line segment, and a second solid line segment. The ordered point set of the lane line includes the points (one or more points) of the first solid line segment, the points (one or more points) of the dotted line segment, and the points (one or more points) of the second solid line segment. At the same time, the first change point and the second change point are recorded. The first change point is the change point from the first solid line segment to the dotted line segment, and the second change point is the change point from the dotted line segment to the second solid line segment. That is, the ordered point set of each lane line may include at least one of the points on at least one solid line segment and at least one dotted line segment arranged in order. If the lane line has at least one solid line segment and at least one dotted line segment, the lane line object information includes at least one change point corresponding to the lane line. According to the endpoints and the real-virtual change points of each lane line in the first lane line group, the first lane line group can be divided into at least one first sub-lane line group. Specifically, each lane line in the first lane line group is divided into at least one group of sub-lane lines of the same or similar length. For example, the first lane line group of the longitudinal type is divided into multiple segments, each segment is a first sub-lane line group. The sub-lane lines in each first sub-lane line group have the same or similar length. Similarly, the second lane line group is divided into at least one second sub-lane line group. The number of first sub-lane line groups and the number of second sub-lane line groups can be the same or different.

[0101] In some optional embodiments, any lane line group in the first lane line group and the second lane line group can be divided at each virtual-real change point of each lane line, and the lane line group can be divided into a solid line sub-lane line group and a dashed line sub-lane line group.

[0102] In some optional embodiments, the first lane line group and the second lane line group may be segmented in combination with the virtual-real change point and the length threshold, and each lane line group may be segmented into sub-lane line groups whose lengths do not exceed the length threshold. The specific segmentation method is not limited.

[0103] In some optional embodiments, after obtaining the first sub-lane line group and the second sub-lane line group, the lane element information and the driving line element information can be determined using the sub-lane line segments in the first sub-lane line group and the second sub-lane line group as granularity, that is, any two adjacent sub-lane lines in each sub-lane line group form a lane element and a driving line element.

[0104] In the disclosed embodiment, by determining the first lane line group in the first direction and the second lane line group in the second direction, the consistency of the lane line direction in each lane line group can be ensured, so that more abundant directional semantic information can be determined for abstract lane elements and driving line elements, further improving the semantic richness of logical element information. Furthermore, by further segmenting the first lane line group and the second lane line group by virtual-real change points, each lane line group can be fine-grained, so as to provide semantic information of the virtual-real type of the lane line for determining the lane element, further improving the semantic richness of the logical element.

[0105] In some optional embodiments, determining lane element information and driving line element information according to the cluster, stop line object information and intersection boundary information corresponding to each lane line group may include:

[0106] According to the direction angle of the lane lines in each cluster corresponding to each lane line group, the first lane line group in the first direction and the second lane line group in the second direction corresponding to each cluster are determined; according to the endpoints and virtual-real change points of the lane lines in the first lane line group corresponding to each cluster, the first lane line group is divided into at least one first sub-lane line group; according to the endpoints and virtual-real change points of the lane lines in the second lane line group corresponding to each cluster, the second lane line group is divided into at least one second sub-lane line group; according to the stop line object information and the intersection boundary information, the lane line parts extending into the intersection in the first sub-lane line group and the second sub-lane line group are filtered out to obtain the third sub-lane line group corresponding to the first sub-lane line group and the fourth sub-lane line group corresponding to the second sub-lane line group; based on the third sub-lane line group and the fourth sub-lane line group, the lane element information and the driving line element information are determined.

[0107] Among them, the specific operation of determining the first sub-lane line group and the second sub-lane line group can refer to the above-mentioned embodiment, which will not be repeated here. After obtaining the first sub-lane line group and the second sub-lane line group, the lane line part extending into the intersection in the first sub-lane line group and the second sub-lane line group can be filtered out (or cut off) according to the stop line object information and the intersection boundary information in the intersection, so that the third sub-lane line group and the fourth sub-lane line group after the cut off are more consistent with the actual situation of the lane lines around the intersection area, that is, there is no physical lane line object in the intersection. Specifically, the target first sub-lane line group and the target second sub-lane line group close to the stop line can be determined according to the relative position relationship between the first sub-lane line group and the second sub-lane line group and the stop line, and the intersection of the target first sub-lane line group and the target second sub-lane line group and the intersection boundary can be determined according to the target first sub-lane line group and the target second sub-lane line group and the intersection boundary information, and the part of the lane line that exceeds the stop line and extends into the intersection can be cut off according to the intersection to obtain the cut third sub-lane line group and the fourth sub-lane line group, and the first sub-lane line group and the second sub-lane line group that do not need to be cut can be directly used as the third sub-lane line group and the fourth sub-lane line group, so that the third sub-lane line group corresponding to each first sub-lane line group and the fourth sub-lane line group corresponding to each second sub-lane line group can be obtained. Then, based on the third sub-lane line group and the fourth sub-lane line group, the lane element information and the driving line element information required for the local map can be determined.

[0108] In the disclosed embodiment, by cutting off the lane lines that extend beyond the stop line into the intersection, the accuracy and effectiveness of the sub-lane line group can be further improved, thereby improving the accuracy and effectiveness of the logical elements.

[0109] Figure 5 It is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure.

[0110] In some optional embodiments, such as Figure 5 As shown, step 2330 of determining the logical element information of the local map according to the lane element information and the driving line element information may include:

[0111] Step 23310, determine the lane segment element information based on the lane element information.

[0112] Among them, each lane segment element is a regional element composed of lane elements of a region. That is, the lane segment element information includes one or more adjacent lane element information that meets the preset conditions. The preset conditions can be set according to the content included in the lane element information. For example, if the lane element information is obtained based on the above steps 23210 to 23240, the preset condition can be the lane element information formed by lane lines belonging to the same cluster. If the lane element information is determined based on the above-mentioned first sub-lane line group and the second sub-lane line group, or based on the above-mentioned third sub-lane line group and the fourth sub-lane line group, the preset condition can be the lane element information formed by the sub-lane lines belonging to the same sub-lane line group, that is, the lane element information with the same driving direction and adjacent in sequence. For example, the area where each lane element corresponding to each sub-lane line group (for example, each first sub-lane line group, each second sub-lane line group, or each third sub-lane line group, each fourth sub-lane line group) is located is abstracted as a lane segment element.

[0113] In some optional embodiments, at least one lane segment element information may be determined based on the association relationship between lane elements in the lane element information. Each lane segment element information may include one or more of the following information: a lane segment element identifier, a bounding box, a center point, a direction, a direction of a lane line included in the lane segment element.

[0114] Step 23320, determine the logical element information of the local map based on the lane element information, the driving line element information and the lane segment element information.

[0115] In some optional embodiments, the obtained lane element information, driving line element information and lane segment element information can be determined as the logical element information required for the local map. Alternatively, based on these element information, other semantic types of element information can be further abstracted, for example, in the intersection scenario, the virtual lane element information, virtual lane segment element information, intersection element information, etc. in the intersection are determined, and used together with the lane element information, driving line element information and lane segment element information to determine the logical element information of the local map.

[0116] The disclosed embodiment further abstracts lane segment element information based on lane element information, further improving the semantic richness of logical elements, thereby improving the semantic richness of the local map.

[0117] Figure 6 It is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure.

[0118] In some optional embodiments, such as Figure 6As shown, step 2330 of determining the logical element information of the local map according to the lane element information and the driving line element information may include:

[0119] Step 233a0: If the current scene type where the movable device is located is an intersection scene, the lane segment element information is determined according to the lane element information.

[0120] Among them, the specific operation of determining the lane segment element information according to the lane element information can refer to the above embodiment.

[0121] Step 233b0, determining the number of intersections and the status information of each intersection according to the lane line object information.

[0122] The status information of each intersection may include the location of the intersection, the type of intersection, the stop line in the intersection, the crosswalk, etc. The intersection type may include a crossroad, a T-intersection, a Y-intersection, an X-intersection, etc.

[0123] In some optional embodiments, the number of intersections and the state information of each intersection may be determined based on at least one cluster corresponding to each lane line group, based on the at least one cluster corresponding to the lane line group obtained in step 23230. That is, step 233b0 may specifically include: determining the number of intersections and the state information of each intersection based on the cluster corresponding to each lane line group.

[0124] In some optional embodiments, each lane line in each cluster can be extended, and the intersection between the extension lines of the lane lines in each cluster can be calculated to obtain at least one intersection. In the intersection area, the horizontal lane lines around the intersection and the extension lines of the longitudinal lane lines into the intersection will intersect, and the horizontal lane lines on both sides of the same intersection may also intersect, and the longitudinal lane lines on both sides of the same intersection may also intersect. For example, when the horizontal lane lines on both sides of the intersection are not parallel, the extension lines extending into the intersection may intersect. Based on this, the obtained intersections can be clustered to obtain at least one intersection cluster, and the number of clusters of the intersection cluster can be used as the number of intersections. The center point of the intersection can be determined according to the intersections in each intersection cluster as the location of the intersection. For example, the position mean of each intersection in the intersection cluster is used as the location of the intersection. According to the location of the intersection and the distance relationship between the physical objects such as the stop line and the crosswalk in the physical object information and the location of the intersection, the physical objects such as the stop line and the crosswalk belonging to the intersection are determined.

[0125] Step 233c0, determining intersection element information according to the number of intersections and the status information of each intersection.

[0126] Among them, the intersection element information may include one or more attribute information of the intersection element, and may also include the logical association relationship between the intersection element and other elements. The attribute information of the intersection element may include, for example, the intersection element identification, the location of the intersection center point, the intersection type and other attribute information. The logical association relationship between the intersection element and other elements may include the physical object information such as the stop line and the crosswalk within the intersection element, the logical element information of the virtual lane element, the virtual lane segment element, etc. within the intersection element, the lane segment element information entering the intersection and the lane segment element information exiting the intersection around the intersection element, etc. The virtual lane element can be determined based on the lane elements around the intersection and the drivable directions of the intersection. For example, for the first side of the intersection (such as Figure 1 Any lane (for example, laneA) on the lower side of the intersection (i.e., the side where the vehicle 11 is located) if the lane laneA can go straight to the second side of the intersection (for example Figure 1 If lane A is on the upper side of the intersection, a virtual lane element is determined between the lane element corresponding to lane A and the lane element corresponding to lane B. If lane A can turn right to the third side of the intersection (for example Figure 1 If the lane C is on the right side of the intersection, a virtual lane element is determined between the element corresponding to lane A and the lane element corresponding to lane C in the intersection. The virtual lane segment element is determined based on the virtual lane element. The determination principle is similar to that of the lane segment element determined in the above embodiment. For example, one or more virtual lane elements in the same direction constitute a virtual lane segment element, which will not be described in detail.

[0127] Step 233d0, determining the logical element information of the local map according to the lane element information, the driving line element information, the lane segment element information and the intersection element information.

[0128] Among them, in the intersection scenario, lane element information, driving line element information, lane segment element information and intersection element information are determined as logical element information of the local map.

[0129] In some optional embodiments, when at least one of a virtual lane element, a virtual lane segment element, and a virtual driving line element exists in an intersection element, the lane element information, the driving line element information, the lane segment element information, the intersection element information, and at least one of the virtual lane element, the virtual lane segment element, and the virtual driving line element can be determined as logical element information of the local map.

[0130] In the embodiment of the present disclosure, in an intersection scenario, on the basis of lane element information, driving line element information and lane segment element information, intersection element information is further determined, which can effectively represent the logical association relationship between intersection elements, elements within the intersection, and elements around the intersection, and further enhance the semantic richness of the local map.

[0131] In some optional embodiments, determining the intersection element information according to the number of intersections and the status information of each intersection in step 233c0 may include:

[0132] According to the status information of each intersection and the historical status information of the intersection, determine whether the intersection is in a stable state; according to whether the intersection is in a stable state, determine the intersection element information.

[0133] Among them, the historical state information of the intersection is the state information of the intersection determined at one or more moments before the current moment. The determination method of the historical state information is similar to that of the state information at the current moment, which will not be described in detail here. The state of whether the intersection is stable can indicate whether the identified intersection is credible, that is, whether the intersection is indeed an intersection. If the intersection is stable, the intersection element information corresponding to the intersection can be determined. If the intersection is unstable, it means that the recognition result of the intersection is not credible, and it is not necessary to determine the intersection element information corresponding to the intersection temporarily. The specific method for determining the intersection element information can refer to the aforementioned embodiment, that is, the intersection element information can be determined according to the process of steps 233b0 and 233c0, which will not be described in detail here.

[0134] In some optional embodiments, a sliding window of a preset size may be set to store the intersection state information of multiple consecutive moments (or multiple frames) including the current moment and the historical moment, to determine whether the intersection is stable. For example, if the intersection state information of multiple consecutive frames is consistent, and the number of consecutive frames is greater than a preset frame number threshold, the intersection is determined to be stable, and the intersection element information corresponding to the intersection can be determined.

[0135] In some optional embodiments, when it is necessary to determine the intersection element information of the intersection, the device status information before the movable device enters the intersection can be recorded. The device status information may include the position and posture of the movable device, the lane element where the movable device is currently located, the lane segment element where the movable device is currently located, and other information, and establish an association relationship between the current position of the movable device and the logical element, so as to facilitate navigation and path planning of the movable device.

[0136] The disclosed embodiments can characterize the credibility of an identified intersection by determining whether the intersection is stable. By combining the stability of the intersection and determining intersection feature information when the intersection is stable, the stability and reliability of the intersection feature information can be improved, thereby improving the reliability of the local map.

[0137] In some optional embodiments, determining the intersection element information according to the number of intersections and the status information of each intersection in step 233c0 may include:

[0138] Determine virtual lane element information within the intersection based on the number of intersections, status information of each intersection, and lane segment element information within the first range of each intersection; determine virtual lane segment element information based on the virtual lane element information within the intersection; determine intersection element information based on the virtual lane element information within the intersection and the virtual lane segment element information.

[0139] The first range of any intersection is the area around the intersection, that is, the area outside the intersection. For example, the first range can be a range of a preset distance outward from the intersection boundary box (or bounding box). The lane segment element information within the first range of the intersection can be determined based on the positional relationship between the lane segment element information obtained in step 233a0 and the intersection.

[0140] In some optional embodiments, the virtual lane element information in the intersection can be determined based on the association relationship between the lane segment element information in the first range of the intersection and the intersection. Exemplarily, the entry and exit relationship of each lane segment element relative to the intersection can be determined based on the position of the intersection (the position of the center point of the intersection), the area range of the intersection, the area range of the lane segment elements in the first range around the intersection, the direction of the lane segment elements and the relative position relationship of the center point. The direction of the lane segment element can be determined based on the drivable direction of the lane in the lane segment element. If the drivable direction of the lane in the lane segment element is toward the intersection, it means that the relationship between the lane segment element and the intersection is an entry relationship, that is, the mobile device can enter the intersection from any lane in the lane segment element. If the drivable direction of the lane in the lane segment element is away from the intersection, it means that the relationship between the lane segment element and the intersection is an exit relationship, that is, the mobile device can exit from the intersection to any lane in the lane segment element. According to the lane segment element information within the first range of the intersection and the entry and exit relationship of the intersection, the entry lane element (which may be referred to as the entry lane element) is connected with one or more exit lane elements (which may be referred to as the exit lane elements) to obtain one or more virtual lane element information corresponding to the entry lane element. For any entry lane element, each exit lane element can be traversed to determine whether a virtual lane element can be determined between the entry lane element and the traversed exit lane element. Specifically, it can be determined whether a virtual lane element needs to be determined between the entry lane element and the traversed exit lane element based on the road surface arrow information associated with the entry lane element. The road surface arrow information associated with the entry lane element is the status information of the road surface arrow included in the lane area corresponding to the entry lane element. The status information of the road surface arrow may include the type, position and other information of the road surface arrow. The type of road surface arrow may include left turn, right turn, straight, U-turn and other types. For example, if the road arrow associated with the entry lane element is a straight-ahead arrow, then for an exit lane element that satisfies the straight-ahead condition relative to the entry lane element, a virtual lane element between the entry lane element and the exit lane element can be determined, and an exit lane element that does not satisfy the straight-ahead condition does not need to determine a virtual lane element. For another example, if the road arrow associated with the entry lane element is a left-turn arrow, then for an exit lane element that satisfies the left-turn condition relative to the entry lane element, a virtual lane element between the entry lane element and the exit lane element can be determined.

[0141] In some optional embodiments, after determining the virtual lane element information in the intersection, the virtual lane segment element information in the intersection can be determined based on the virtual lane element information. The virtual lane segment element information is similar to the lane segment element information, that is, the virtual lane element information in the intersection is determined according to the direction type of the virtual lane element (for example, the direction type of the road arrow such as straight, left turn, right turn, U-turn, etc.), and the virtual lane elements of the same direction type constitute the corresponding virtual lane segment elements to obtain the virtual lane segment element information. The specific information content included in the virtual lane segment element information can refer to the lane segment element information, which will not be repeated here. Then, the intersection element information can be determined in combination with the virtual lane element information and the virtual lane segment element information in the intersection. Optionally, on the basis of the virtual lane element information and the virtual lane segment element information, the intersection element information can be further determined in combination with the physical object information such as the stop line and the crosswalk in the intersection. Physical objects such as stop lines and crosswalks can be used as sub-elements or attribute information of intersection elements and recorded in the intersection element information to represent the logical relationship between the intersection and the stop lines and crosswalks.

[0142] In some optional embodiments, the logical element information can be represented as information of a preset data structure, for example, the logical element information of each logical element is represented as a structure having one or more preset fields, and each preset field records a type of information of the logical element. The preset field may also include a field representing the association relationship between different logical elements, and the association relationship may include a predecessor-successor relationship, a left-right association relationship, a parent-child association relationship, etc. The predecessor-successor relationship represents the adjacent relationship of the logical elements in the direction of the lane line, the left-right association relationship represents the adjacent relationship of the logical elements in the direction perpendicular to the lane line, and the parent-child relationship represents the positional inclusion relationship of the logical elements. For example, if the predecessor of lane element B is lane element A, it means that when driving along the drivable direction of the lane, it will first pass through lane element A and then pass through lane element B. If the left lane element of lane element C is lane element D, it means that lane element C and lane element D are two lane elements adjacent to each other on the left and right. If the parent lane element of virtual lane element E is intersection element F, it means that virtual lane element E is within the area of ​​intersection element F.

[0143] The disclosed embodiment of the present invention fine-grains the intersection feature information into virtual lane feature information and virtual lane segment feature information within the intersection, which can effectively characterize the logical association relationship of the lane feature information around the intersection within the intersection, and further enhance the semantic richness of the local map.

[0144] Figure 7 It is a flowchart of a method for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure.

[0145] In some optional embodiments, based on any of the above embodiments, Figure 7 As shown, step 240 of constructing a local map of a preset range of the current position of the movable device according to the logical element information may include:

[0146] Step 2410, perform element tracking on the logical element information to obtain the matching relationship between the logical element information and the historical logical element information.

[0147] The logical element information is the logical element information obtained by the processing flow of constructing the local map at the current moment, and the historical logical element information is the logical element information obtained by the processing flow at the historical moment before the current moment (such as the previous moment). The matching relationship between the logical element information at the current moment and the historical logical element information indicates whether the logical element at the current moment and the historical logical element are the same element.

[0148] In some optional embodiments, any feasible target tracking method can be used to track the elements of the logical element information to obtain the matching relationship between the logical element information and the historical logical element information. For example, the logical element information can be tracked based on the Kalman filter to obtain the matching relationship between the logical element information and the historical logical element information. Specifically, the state of the historical logical element information can be estimated based on the Kalman filter to obtain the estimated amount corresponding to the historical logical element information, the logical element information at the current moment can be used as the observed amount, the similarity between the estimated amount and the observed amount can be calculated, and the matching result between the logical element information at the current moment and each historical logical element information can be obtained based on the similarity, the observed amount can be updated based on the matching result, and the updated observed amount after successful tracking can be obtained, and then the matching relationship between the logical element information and the historical logical element information can be obtained based on the successfully tracked observed amount.

[0149] Step 2420: Update the element identifier in the logical element information according to the matching relationship.

[0150] Among them, after obtaining the matching relationship between the logical element information and the historical logical element information, the element identifier in the logical element information can be updated to the element identifier of the historical logical element information matching the logical element information to indicate that the logical element information and the historical logical element information belong to the element information of the same logical element.

[0151] In some optional embodiments, in the case of element tracking based on the Kalman filter, if the state of the logical element information as the observation quantity is updated, the information related to the observation quantity in the logical element information can also be updated according to the updated observation quantity. For example, the position, direction and other information of the logical element in the logical element information can be updated according to the updated observation quantity.

[0152] Step 2430, determining a local map based on the association relationship between elements in the logical element information.

[0153] The logical element information may include one or more of the lane element information, lane segment element information, driving line element information, intersection element information, virtual lane element information within the intersection, virtual lane segment element within the intersection, virtual driving line element information within the intersection, etc. The local map may be determined based on the association between the elements in the logical element information.

[0154] In some optional embodiments, other element information (including element identifiers of other elements) having a corresponding association relationship with the element may be recorded in a field indicating an association relationship in the logical element information corresponding to each element. For example, one or more of the element's predecessor element identifier, successor element identifier, element identifier associated with the left side (such as adjacent), element identifier associated with the right side, parent element identifier, and child element identifier may be recorded in the structure of the element. The specific association relationship field may be set according to the specific type of the logical element.

[0155] In some optional embodiments, unstable elements can also be filtered out through element tracking to ensure the stability and reliability of logical elements.

[0156] The disclosed embodiment can determine a unique element identifier for a logical element through element tracking, merge the logical element at the current moment with the same logical element at the historical moment, and ensure the uniqueness and reliability of the logical element.

[0157] In some optional embodiments, Figure 8 FIG. 1 is a flowchart of a method for constructing a local map during driving of a mobile device provided by an exemplary embodiment of the present disclosure. Figure 8 As shown, the process of the method may include the following steps:

[0158] Step 310: Collect environment images.

[0159] Step 320: Perform perception processing on the environmental image to obtain a first perception result.

[0160] Step 330: Determine the physical object information. That is, determine the physical object information in the local coordinate system where the movable device is located according to the first perception result.

[0161] Step 340 : Roughly group the lane lines. Go to step 350 and step 410 .

[0162] Specifically, the lane line rough grouping includes: determining the lane line object information in the environment where the mobile device is located according to the physical object information; determining the preset direction type of each lane line according to the lane line object information; grouping the lane lines in the lane line object information based on the preset direction type of each lane line to obtain at least one lane line group; for each lane line group in at least one lane line group, clustering the lane line group to obtain at least one cluster corresponding to the lane line group; and taking the cluster corresponding to each lane line group as the lane line rough grouping result. Alternatively, determining the direction angle of each lane line according to the lane line object information; clustering the lane lines based on the direction angle of each lane line to obtain at least one cluster of the preset direction type, taking each cluster as a lane line group, and obtaining at least one lane line group; for each lane line group in at least one lane line group, clustering the lane line group to obtain at least one cluster corresponding to the lane line group; and taking the cluster corresponding to each lane line group as the lane line rough grouping result.

[0163] In some optional examples, Fig. 9 FIG. 1 is an exemplary schematic diagram of a lane line rough grouping result provided by an exemplary embodiment of the present disclosure. Fig. 9 As shown, for non-intersection scenarios, if there is only one continuous road within the perception range, and the first perception result includes four longitudinal lane lines 13, the lane line coarse grouping result may include a cluster 17 corresponding to one lane line group (i.e., four longitudinal lane lines 13 constitute one lane line group).

[0164] In some optional examples, Fig.10 FIG. 1 is another exemplary schematic diagram of a lane line rough grouping result provided by an exemplary embodiment of the present disclosure. Fig.10 As shown, for the intersection scene, taking an intersection J in the figure as an example, the first perception result includes 12 longitudinal lane lines 131 and 12 transverse lane lines 132. The lane line coarse grouping result may include two clusters corresponding to each lane line group in the two lane line groups (i.e., the transverse lane line group and the longitudinal lane line group). The transverse lane line group includes 12 transverse lane lines 132, and the longitudinal lane line group includes 12 longitudinal lane lines 131. The transverse lane line group corresponds to two clusters, such as clusters 171 and cluster 172 in the figure. The longitudinal lane line group corresponds to two clusters, such as clusters 173 and cluster 174 in the figure.

[0165] Above Fig. 9 and Fig.10 This is only an exemplary description and is not intended to limit the embodiments of the present disclosure.

[0166] Step 350, lane line subdivision grouping. That is, according to the direction angle of the lane line in each cluster corresponding to each lane line group obtained in step 340, determine the first lane line group in the first direction and the second lane line group in the second direction corresponding to each cluster.

[0167] Step 360: Lane line clipping.

[0168] That is, the first lane line group in the first direction and the second lane line group in the second direction corresponding to each cluster obtained in step 350 are divided. Specifically, according to the endpoints and virtual-real change points of the lane lines in the first lane line group corresponding to each cluster, the first lane line group is divided into at least one first sub-lane line group; according to the endpoints and virtual-real change points of the lane lines in the second lane line group corresponding to each cluster, the second lane line group is divided into at least one second sub-lane line group.

[0169] In some optional examples, Fig.11 Schematic diagram of lane line subdivision provided by an exemplary embodiment of the present disclosure. Fig.11 As shown, for cluster 171, if among the six transverse lane lines 132, the lower three transverse lane lines 132 are in the first direction f1 (the direction from left to right in the figure), and the upper three transverse lane lines 132 are in the second direction f2 (the direction from right to left in the figure), then cluster 171 is subdivided into a first lane line group 1711 (including the lower three transverse lane lines 132) in the first direction f1 and a second lane line group 1712 (including the upper three transverse lane lines 132) in the second direction f2. The first direction and the second direction corresponding to different clusters may be different. For example, in cluster 174, the first direction f1 is the direction from bottom to top in the figure, and the second direction f2 is the direction from top to bottom. Cluster 174 can be subdivided into a first lane line group 1741 in the first direction f1 and a second lane line group 1742 in the second direction f2. Then, each subdivided first lane line group and each second lane line are segmented to obtain at least one first sub-lane line group corresponding to each first lane line group and at least one second sub-lane line group corresponding to each second lane line group. Fig.12 Schematic diagram of lane line cutting provided by an exemplary embodiment of the present disclosure. Fig.12 As shown, in Fig.11 On the basis of this, taking the first lane line group 1741 as an example, according to the endpoints P1 and P2 and the virtual-real change points (such as b1 and b2 in the figure) of each lane line in the first lane line group 1741, the first lane line group 1741 is divided into three first sub-lane line groups, such as the first sub-lane line group 17411, the first sub-lane line group 17412 and the first sub-lane line group 17413 in the figure.

[0170] In practical applications, lane line clipping may also include cutting off the portion of the first sub-lane line group and the second sub-lane line group that extends into the intersection, and using the first sub-lane line group after the cutting off as the first sub-lane line group, and using the second sub-lane line group after the cutting off as the second sub-lane line group for subsequent processing.

[0171] Step 370, determine lane element information and driving line element information. That is, determine lane element information and driving line element information based on each first sub-lane line group and each second sub-lane line group obtained in step 360. Two adjacent lane lines in each sub-lane line group form a lane element and a driving line element.

[0172] Step 380: Determine lane segment element information, that is, determine lane segment element information according to lane element information.

[0173] Step 410 , judging the current scene type according to the lane line rough grouping result. If the current scene type is not an intersection scene (ie, a non-intersection scene), proceed to step 420 ; if the current scene type is an intersection scene, proceed to step 510 .

[0174] Step 420 , determining the logical element information. That is, according to the lane element information and the driving line element information obtained in step 370 , and the lane segment element information obtained in step 380 , determining the logical element information required for the local map. Go to step 560 .

[0175] Step 510, intersection information extraction. That is, in the case of an intersection scenario, the intersection information is extracted according to the physical object information obtained in step 330. Specifically, the number of intersections and the status information of each intersection can be determined according to the lane line object information.

[0176] Step 520, determining virtual lane element information in the intersection, that is, determining virtual lane element information in the intersection according to the number of intersections, the state information of each intersection, and the lane segment element information located within the first range around each intersection in the lane segment element information obtained in step 380.

[0177] Step 530, determining virtual lane segment element information in the intersection, that is, determining virtual lane segment element information according to the virtual lane element information in the intersection.

[0178] Step 540, determining the intersection element information, that is, determining the intersection element information according to the virtual lane element information and the virtual lane segment element information in the intersection.

[0179] Step 550, determine the logical element information. That is, according to the lane element information and driving line element information obtained in step 370, the lane segment element information obtained in step 380, the virtual lane element information obtained in step 520, the virtual lane segment element information obtained in step 530, and the intersection element information obtained in step 540, determine the logical element information required for the local map. Go to step 560.

[0180] Step 560: element tracking, that is, element tracking is performed on the logical element information to obtain a matching relationship between the logical element information and the historical logical element information.

[0181] Step 570: logical element update, that is, update the element identifier in the logical element information according to the matching relationship, and may also update other related information.

[0182] In some optional embodiments, on the basis of updating the element identification, for each lane element information, the lane element's predecessor element identification, successor element identification, left-side associated lane element identification, right-side associated lane element identification, lane segment element identification to which the lane element belongs, etc. may be recorded in the lane element information. For each lane segment element information, the lane segment element's predecessor lane segment element identification, successor lane segment element identification, left-side associated lane segment element identification, right-side associated lane segment element identification, and internal sub-element identification may be recorded in the lane segment element information. The internal sub-element identification may include an internal lane element identification, an arrow element identification, etc. For each virtual lane element information and each virtual lane segment element information, the content to be associated is similar to the above-mentioned lane element information and lane segment element information. It is also possible to add an association relationship field different from the lane element information and lane segment element information in the virtual lane element information and the virtual lane segment element information according to different situations in the intersection and non-intersection scenes, without specific limitation. For each intersection element information, the virtual lane element identifier and virtual lane segment element identifier inside the intersection element, the entry lane segment element identifier and exit lane segment element identifier associated with the intersection element outside the intersection element, etc. can be recorded in the intersection element information.

[0183] Step 580, element association. That is, according to the association relationship between elements, each logical element is associated with physical elements such as road arrows, signboards, traffic lights, stop lines, crosswalks, construction areas, cones, etc., and the identification and semantic information of each associated element are recorded in the logical element information. Specifically, an association relationship field related to the physical element can be set in the logical element information, and the identification, semantic information, etc. of each element can be assigned to the corresponding field to represent the association relationship between the logical element and the physical element in the local map. The semantic information of an element refers to the physical meaning represented by the element. For example, the semantic information of an arrow element can include semantic information such as left turn, straight go, right turn, U-turn, etc. The details will not be repeated one by one.

[0184] Step 590, determining a local map. That is, determining a local map based on the associated logical element information of step 580. Specifically, the associated logical element information can represent the logical association relationship of the roads, and the associated logical element information can be determined as a local map, or the associated logical element information can be visualized to obtain a visualized local map for navigation of the mobile device.

[0185] In some optional embodiments, the logical element information corresponding to each logical element can be stored in a designated storage space in a certain order, and the association relationship between the elements is established through the element identifier. When a local map is needed, the logical element information of the associated element can be read from the storage space according to the associated element identifier recorded in the association relationship field in the logical element information. For example, the data structure of the logical element information of each logical element is a structure, and the structure of each logical element is stored in a certain order, and the mapping relationship between the element identifier and the storage address can be recorded, so that when navigating based on the local map, the storage address of the structure can be determined according to the element identifier, and the relevant information of the logical element recorded in the structure can be obtained. The specific access method of the logical element information is not limited.

[0186] The disclosed embodiments can construct a local logical map online in real time in the absence of a high-precision map, relying only on the perceived and reconstructed physical element information. As a local map, the local map can include, on the basis of the semantic information of the physical elements, lane elements, lane segment elements, driving line elements, virtual lane elements in intersections, virtual lane segment elements, virtual driving line elements, intersection elements and other logical layer element information that can characterize the logical association relationship of roads, thereby greatly improving the semantic richness of the local map, providing accurate road logical relationships for the navigation of mobile devices, and enabling the constructed local map to meet the functional requirements of lane changing, traffic light start and stop recognition at intersections, U-turns and other functions in the absence of a high-precision map.

[0187] In some optional embodiments, Fig.13FIG. 1 is a visualization diagram of a logic element provided by an exemplary embodiment of the present disclosure. Fig.13 As shown, taking a non-intersection scene as an example, the logical elements may include a lane element 61, a driving line element 62, and a lane segment element 63. Each lane element 61 corresponds to a segment of a lane. Each lane element 61 may include one or more of a sub-lane line 611 constituting a lane, an arrow 612, and a driving line element 62 on the lane. Each lane segment element 63 corresponds to a segment area where a group of lanes in the same direction are located, such as the area surrounded by the black frame in the figure. The lane segment element 63 may include, for example, each lane element 61 in the segment area.

[0188] In some optional embodiments, Fig.14 FIG. 1 is a visualization diagram of a logic element provided by another exemplary embodiment of the present disclosure. Fig.14 As shown, taking the intersection scene as an example, the logical element may include an intersection element 64 on the basis of the lane element 61, the driving line element 62, and the lane segment element 63 (not shown in the figure). The intersection element 64 may include a virtual lane element 641 and a virtual lane segment element 642 in the intersection. The intersection element 64 may also include physical elements such as a stop line 65 and a zebra crossing (i.e., a pedestrian crossing) 66 in the intersection. For the virtual lane element 641, only some longitudinal virtual lane elements are shown in the figure. In actual applications, the intersection element 64 may also include a transverse virtual lane element (e.g., a virtual lane element from the left side of the intersection to the right side of the intersection or from the right side to the left side), a turning virtual lane element, etc. The turning virtual lane element may include, for example, a left-turn virtual lane element, a right-turn virtual lane element, etc. In actual applications, the intersection element 64 may also include a virtual lane driving line element and other related elements, or the virtual lane segment element 642 may include a virtual lane driving line element and other related elements. It is not specifically limited.

[0189] The above-mentioned embodiments of the present disclosure may be implemented separately or in any combination without conflict. The specific configuration may be based on actual needs and the present disclosure does not limit this.

[0190] Any method for constructing a local map during the driving of a mobile device provided in the embodiments of the present disclosure can be executed by any appropriate electronic device with data processing capabilities, including but not limited to: electronic devices such as terminal devices and servers. Alternatively, any method for constructing a local map during the driving of a mobile device provided in the embodiments of the present disclosure can be executed by a processor, such as the processor executing any method for constructing a local map during the driving of a mobile device mentioned in the embodiments of the present disclosure by calling corresponding instructions stored in a memory. This will not be described in detail below.

[0191] Exemplary Devices

[0192] Fig.15 FIG. 1 is a schematic diagram of a structure of an apparatus for constructing a local map during the driving of a mobile device provided by an exemplary embodiment of the present disclosure. The apparatus of this embodiment can be used to implement the corresponding method embodiments of the present disclosure, such as Fig.15 The device shown may include: a first processing module 71 , a second processing module 72 , a third processing module 73 and a fourth processing module 74 .

[0193] The first processing module 71 is used to determine a first perception result based on an environment image of the movable device during driving.

[0194] The second processing module 72 is used to determine the physical object information in the local coordinate system where the movable device is located according to the first perception result.

[0195] The third processing module 73 is used to determine the logical element information of the local map to be constructed according to the physical object information; the logical element information is used to characterize the logical association relationship of the roads.

[0196] The fourth processing module 74 is used to construct a local map of a preset range of the current position of the movable device according to the logical element information.

[0197] In some optional embodiments, the first processing module 71 may be specifically configured to determine a first perception result based on an environmental image and at least one other sensor data.

[0198] In some optional embodiments, the logical element information may also be used to represent the road semantic information.

[0199] In some optional embodiments, the fourth processing module 74 may be specifically used to construct a local map of a preset range of the current position of the movable device based on the physical object information and the logical element information.

[0200] Fig.16 It is a structural schematic diagram of an apparatus for constructing a local map during driving of a movable device provided by another exemplary embodiment of the present disclosure.

[0201] In some optional embodiments, in the above Fig.15 Based on the embodiment shown, Fig.16 As shown, the third processing module 73 may include: a first processing unit 731 , a second processing unit 732 and a third processing unit 733 .

[0202] The first processing unit 731 is used to determine lane line object information in the environment where the movable device is located according to the physical object information.

[0203] The second processing unit 732 is used to determine the lane element information and driving line element information required for the local map according to the lane line object information.

[0204] The third processing unit 733 is used to determine the logical element information of the local map according to the lane element information and the driving line element information.

[0205] In some optional embodiments, the second processing unit 732 may be specifically configured to:

[0206] According to the lane line object information, the preset direction type of each lane line is determined; based on the preset direction type of each lane line, the lane lines in the lane line object information are grouped to obtain at least one lane line group; for each lane line group in the at least one lane line group, the lane line group is clustered to obtain at least one cluster corresponding to the lane line group; based on the cluster corresponding to each lane line group, the lane element information and the driving line element information are determined.

[0207] In some optional embodiments, the second processing unit 732 can be specifically used to: determine the direction angle of each lane line based on the lane line object information; cluster the lane lines according to the direction angle of each lane line to obtain at least one cluster; determine the preset direction type of each cluster; use each cluster as a lane line group of a preset direction type; for each lane line group, cluster the lane line group to obtain at least one cluster corresponding to the lane line group; determine the lane element information and driving line element information based on the cluster corresponding to each lane line group.

[0208] In some optional embodiments, the second processing unit 732 can be specifically used to: for each lane line group in at least one lane line group, determine the center point of each lane line in the lane line group; based on the center point of each lane line, cluster the lane line group to obtain at least one cluster corresponding to the lane line group.

[0209] In some optional embodiments, the second processing unit 732 may be specifically configured to:

[0210] Based on the clusters corresponding to each lane line group, the current scene type where the movable device is located is determined; based on the current scene type and the clusters corresponding to each lane line group, lane element information and driving line element information are determined.

[0211] In some optional embodiments, the second processing unit 732 may be specifically configured to:

[0212] If the current scene type is an intersection scene, the stop line object information and intersection boundary information within the intersection are determined based on the physical object information; the lane element information and driving line element information are determined based on the clustering, stop line object information and intersection boundary information corresponding to each lane line group.

[0213] In some optional embodiments, the second processing unit 732 may be specifically configured to:

[0214] According to the direction angle of the lane lines in each cluster corresponding to each lane line group, the first lane line group in the first direction and the second lane line group in the second direction corresponding to each cluster are determined; according to the endpoints and virtual-real change points of the lane lines in the first lane line group corresponding to each cluster, the first lane line group is divided into at least one first sub-lane line group; according to the endpoints and virtual-real change points of the lane lines in the second lane line group corresponding to each cluster, the second lane line group is divided into at least one second sub-lane line group; based on the first sub-lane line group and the second sub-lane line group, the lane element information and the driving line element information are determined.

[0215] In some optional embodiments, the third processing unit 733 may be specifically configured to:

[0216] The lane segment element information is determined according to the lane element information; and the logical element information of the local map is determined according to the lane element information, the driving line element information and the lane segment element information.

[0217] In some optional embodiments, the third processing unit 733 may be specifically configured to:

[0218] If the current scene type where the mobile device is located is an intersection scene, determine the lane segment element information based on the lane element information; determine the number of intersections and the status information of each intersection based on the lane line object information; determine the intersection element information based on the number of intersections and the status information of each intersection; determine the logical element information of the local map based on the lane element information, driving line element information, lane segment element information and intersection element information.

[0219] In some optional embodiments, the third processing unit 733 may be specifically configured to:

[0220] According to the status information of each intersection and the historical status information of the intersection, determine whether the intersection is in a stable state; according to whether the intersection is in a stable state, determine the intersection element information.

[0221] In some optional embodiments, the third processing unit 733 may be specifically configured to:

[0222] Determine virtual lane element information within the intersection based on the number of intersections, status information of each intersection, and lane segment element information within the first range of each intersection; determine virtual lane segment element information based on the virtual lane element information within the intersection; determine intersection element information based on the virtual lane element information within the intersection and the virtual lane segment element information.

[0223] Fig.17It is a structural schematic diagram of an apparatus for constructing a local map during driving of a movable device provided by yet another exemplary embodiment of the present disclosure.

[0224] In some optional embodiments, based on any of the above embodiments, Fig.17 As shown, the fourth processing module 74 may include: a fourth processing unit 741 , a fifth processing unit 742 and a sixth processing unit 743 .

[0225] The fourth processing unit 741 is used to track the elements of the logical element information and obtain the matching relationship between the logical element information and the historical logical element information.

[0226] The fifth processing unit 742 is used to update the element identifier in the logical element information according to the matching relationship.

[0227] The sixth processing unit 743 is used to determine the local map based on the association relationship between the elements in the logical element information.

[0228] The specific operations of each module and each unit in the exemplary embodiment of the present device can refer to the specific contents of the corresponding steps in the above exemplary method part, which will not be repeated here.

[0229] The beneficial technical effects corresponding to the exemplary embodiment of the present device can be found in the corresponding beneficial technical effects of the above exemplary method section, which will not be repeated here.

[0230] Exemplary Electronic Devices

[0231] Fig.18 It is a structural diagram of an electronic device provided by an embodiment of the present disclosure, including at least one processor 91 and a memory 92.

[0232] The processor 91 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 90 to perform desired functions.

[0233] The memory 92 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 91 may execute one or more computer program instructions to implement the methods and / or other desired functions of the various embodiments of the present disclosure described above.

[0234] In one example, the electronic device 90 may further include: an input device 93 and an output device 94 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0235] The input device 93 may also include, for example, a touch screen, a microphone, various sensors, and the like. Sensors may include, for example, image sensors (such as cameras, cameras, etc.), laser radars, millimeter-wave radars, ultrasonic radars, positioning sensors, pressure sensors, air quality sensors, temperature sensors, and the like. Image sensors, laser radars, millimeter-wave radars, ultrasonic radars, and the like may be used to sense the surrounding environment, that is, to detect dynamic and static objects in the surrounding environment. Dynamic and static objects may include, for example, static objects such as lane lines, curbs, arrows, signs, trees, and buildings, as well as dynamic objects such as surrounding vehicles, pedestrians, and cyclists. Positioning sensors are used to realize the positioning of movable devices (such as self-vehicles, robots, etc.) where electronic devices are located. Positioning sensors may include, for example, inertial measurement units (Inertial Measurement Unit, referred to as: IMU), global positioning systems (Global Positioning System, referred to as: GPS), and the like. Pressure sensors may be used to detect seat pressure. Temperature sensors may be used to detect the temperature in the vehicle cabin. Air quality sensors may be used to detect the air quality in the vehicle cabin.

[0236] The output device 94 can output various information to the outside, and may include, for example, a display, a speaker, a communication network and a remote output device connected thereto, and the like.

[0237] Of course, to simplify, Fig.18 Only some of the components related to the present disclosure in the electronic device 90 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 90 may also include any other appropriate components.

[0238] Exemplary computer program products and computer-readable storage media

[0239] In addition to the above methods and devices, embodiments of the present disclosure may also provide a computer program product, including computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the methods of various embodiments of the present disclosure described in the above “Exemplary Methods” section.

[0240] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0241] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the processor executes the steps in the method of various embodiments of the present disclosure described in the above “Exemplary Method” section.

[0242] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium is, for example, but not limited to, a system, device or device including electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0243] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and cannot be considered as necessary for each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, rather than limitation, and the above details do not limit the present disclosure to being implemented by adopting the above specific details.

[0244] Those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present disclosure claims and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A method for constructing a local map during the driving of a movable device, comprising: Determine a first perception result based on an image of the environment of the movable device during driving; Determine, according to the first perception result, physical object information in the local coordinate system where the movable device is located; Determine logical element information of the local map to be constructed according to the physical object information; the logical element information is used to characterize the logical association relationship of the roads; A local map of a preset range of the current position of the movable device is constructed according to the logical element information.

2. The method according to claim 1, wherein: The step of determining the logical element information of the local map to be constructed according to the physical object information includes: Determine lane line object information in the environment where the movable device is located according to the physical object information; Determining lane element information and driving line element information required for the local map according to the lane line object information; The logical element information of the local map is determined according to the lane element information and the driving line element information.

3. The method according to claim 2, wherein: The determining, according to the lane line object information, lane element information and driving line element information required by the local map includes: Determining a preset direction type of each lane line according to the lane line object information; Based on the preset direction type of each lane line, grouping the lane lines in the lane line object information to obtain at least one lane line group; For each lane line group in the at least one lane line group, clustering the lane line group to obtain at least one cluster corresponding to the lane line group; The lane element information and the driving line element information are determined based on the cluster corresponding to each lane line group.

4. The method according to claim 3, wherein: The step of clustering each lane line group in the at least one lane line group to obtain at least one cluster corresponding to the lane line group includes: For each lane line group in the at least one lane line group, determining a center point of each lane line in the lane line group; Based on the center point of each lane line, the lane line group is clustered to obtain the at least one cluster corresponding to the lane line group.

5. The method according to claim 3, wherein: The determining the lane element information and the driving line element information based on the cluster corresponding to each lane line group includes: Determining a current scene type where the movable device is located based on the clusters corresponding to each lane line group; The lane element information and the driving line element information are determined based on the current scene type and the cluster corresponding to each lane line group.

6. The method according to claim 3, wherein: The determining the lane element information and the driving line element information based on the cluster corresponding to each lane line group includes: Determine a first lane line group in a first direction and a second lane line group in a second direction corresponding to each cluster according to the direction angle of the lane line in each cluster corresponding to each lane line group; Divide the first lane line group into at least one first sub-lane line group according to the endpoints and the virtual-real change points of the lane lines in the first lane line group corresponding to each cluster; Divide the second lane line group into at least one second lane line sub-group according to the endpoints and the virtual-real change points of the lane lines in the second lane line group corresponding to each cluster; The lane element information and the driving line element information are determined based on the first sub-lane line group and the second sub-lane line group.

7. The method according to claim 2, wherein: The determining the logical element information of the local map according to the lane element information and the driving line element information includes: Determining lane segment element information according to the lane element information; The logical element information of the local map is determined according to the lane element information, the driving line element information and the lane segment element information.

8. The method according to claim 5, wherein: The determining the lane element information and the driving line element information based on the current scene type and the cluster corresponding to each lane line group includes: If the current scene type is an intersection scene, determining stop line object information and intersection boundary information within the intersection according to the physical object information; The lane element information and the driving line element information are determined according to the cluster corresponding to each lane line group, the stop line object information and the intersection boundary information.

9. The method according to claim 2, wherein: The determining the logical element information of the local map according to the lane element information and the driving line element information includes: If the current scene type where the movable device is located is an intersection scene, determining lane segment element information according to the lane element information; Determine the number of intersections and status information of each intersection according to the lane line object information; Determining intersection element information according to the number of intersections and status information of each intersection; The logical element information of the local map is determined according to the lane element information, the driving line element information, the lane segment element information and the intersection element information.

10. The method according to any one of claims 1 to 9, wherein: The step of constructing a local map of a preset range of the current position of the movable device according to the logical element information includes: Performing element tracking on the logical element information to obtain a matching relationship between the logical element information and historical logical element information; According to the matching relationship, updating the element identifier in the logical element information; The local map is determined based on the association relationship between the elements in the logical element information.

11. A device for constructing a local map during the driving of a movable device, comprising: A first processing module, configured to determine a first perception result based on an environmental image of the movable device during driving; A second processing module, configured to determine physical object information in a local coordinate system where the movable device is located according to the first sensing result; A third processing module is used to determine the logical element information of the local map to be constructed according to the physical object information; the logical element information is used to characterize the logical association relationship of the roads; The fourth processing module is used to construct a local map of a preset range of the current position of the movable device according to the logical element information.

12. A computer-readable storage medium storing a computer program, wherein the computer program is used to execute the method according to any one of claims 1 to 10.

13. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1-10.

Images