3D precision pedestrian map production system and method

Abstract

A system and method for producing a 3D precision pedestrian map are disclosed. A 3D precision pedestrian map producing system according to one embodiment of the present invention includes: a data processing unit that obtains processed precision pedestrian map data by performing processing on public data, including height application and normalization processing according to preset processing standards; a data modeling unit that obtains modeling data by performing modeling processing on processed precision pedestrian map data according to preset modeling standards, including terrain group modeling, building modeling and placement, and pedestrian obstacle modeling and placement; and a data providing unit that provides modeling data, and generates and provides a 3D precision pedestrian map by converting the modeling data according to preset data conversion standards.

Description

Technology Field

[0001] The disclosed embodiments relate to a 3D precision walking map production system and method. Background Technology

[0002] With the operation of autonomous wheelchairs and delivery robots becoming a reality in preparation for a super-aged society, there is a growing demand for the creation of separate, high-precision maps for pedestrian movement spaces.

[0003] Meanwhile, pedestrian movement spaces such as sidewalks, pedestrian-only roads, and side roads need to be differentiated from precise road maps in that, compared to vehicle movement spaces, the length of the sections to be surveyed is long, the spatial configuration is irregular, and the operation of the MMS (Mobile Mapping System) is difficult, while the speed of moving objects is relatively slow, so positioning errors at the level of precise road maps are not required.

[0004] Accordingly, unlike the method of producing precise road maps, a method for producing precise pedestrian maps is being sought that enables rapid analysis processing of extensive areas at a low cost. Prior art literature

[0006] Republic of Korea Published Patent Application No. 10-2023-0075741 (May 31, 2023) The problem to be solved

[0007] The disclosed embodiments aim to provide a 3D precision pedestrian map production system and method for constructing a 3D precision pedestrian map capable of precisely reflecting movement characteristics at a microscopic scale at the pedestrian level.

[0008] In addition, the disclosed embodiments aim to provide a 3D precision pedestrian map production system and method for efficiently representing the complex structure of a city from a pedestrian's perspective.

[0009] In addition, the disclosed embodiments aim to provide a 3D precision pedestrian map production system and method capable of producing a high-level precision pedestrian map by utilizing public data without direct urban surveying and complex vector mapping processes. means of solving the problem

[0011] A 3D precision pedestrian map production system according to one embodiment includes: a data processing unit that obtains processed precision pedestrian map data by performing processing on public data, including height application and normalization processing, according to preset processing standards; a data modeling unit that obtains modeling data by performing modeling processing on the processed precision pedestrian map data according to preset modeling standards, including terrain group modeling, building modeling and placement, and pedestrian obstacle modeling and placement; and a data providing unit that provides the modeling data, and generates and provides a 3D precision pedestrian map by converting the modeling data according to preset data conversion standards.

[0012] The data processing unit above can predefine the 3D terrain of the 3D precise pedestrian map, including basic pedestrian environment objects such as roads, sidewalks, crosswalks, overpasses, parcels, traffic islands, and water spaces.

[0013] The above data processing unit can obtain the processed precision pedestrian map data through the processing of the above public data, which includes processing terrain-related data, processing building data, and processing pedestrian obstacle data.

[0014] The data modeling unit can obtain an optimized terrain (TerrainOpt) using basic terrain data among the processed precision walking map data when modeling the terrain group, obtain water space data through water space data modeling processing, and obtain a terrain group object, which is the modeling data, by processing a pre-set detail element modification modeling using at least one of the optimized terrain, the water space data, and additional terrain data.

[0015] The data modeling unit can, when performing the building modeling and placement, extract a Z value according to a pre-set first modeling standard using processed building data among the processed precision pedestrian map data, classify a list of Z value data containing at least one of the extracted Z values ​​to extract a minimum Z value and a maximum Z value, define a landscape height using these values, generate a building basic height by applying a pre-set average height per floor to the basic attribute of the number of floors among the processed building data, generate a building height using the landscape height and the building basic height, extrude BuildingSrf data by the building height, and complete the building arrangement by moving and placing the extruded data according to a pre-set second modeling standard.

[0016] The data modeling unit above can perform processing such as placing pedestrian obstacle element data among the processed precision pedestrian map data onto the terrain according to a preset automated placement algorithm when modeling and placing the pedestrian obstacle element data, applying FlipMatrix texture mapping, aligning terrain Z coordinates based on MeshRay, and placing it so as to be perpendicular to the road direction of travel.

[0017] The above data providing unit can convert the above modeling data according to the above data conversion standard, which includes at least one of spatial error optimization and 3D coordinate range optimization.

[0018] A method for producing a 3D precise pedestrian map according to one embodiment comprises, in a method performed by a 3D precise pedestrian map production system, a step of obtaining processed precise pedestrian map data by performing processing on public data, including height application and normalization processing according to preset processing standards; a step of obtaining modeling data by performing modeling processing on the processed precise pedestrian map data according to preset modeling standards, including terrain group modeling, building modeling and placement, and pedestrian obstacle modeling and placement; and a step of providing the modeling data, wherein the modeling data is converted according to preset data conversion standards to generate and provide a 3D precise pedestrian map.

[0019] The above method for producing a 3D precise pedestrian map may further include the step of pre-defining the 3D terrain of the 3D precise pedestrian map, including basic pedestrian environment objects such as roads, sidewalks, crosswalks, overpasses, parcels, traffic islands, and water spaces.

[0020] The above 3D precision pedestrian map production method can acquire the above precision pedestrian map data through the processing of the public data, including processing of terrain-related data, processing of building data, and processing of pedestrian obstacle data, in the step of acquiring the above precision pedestrian map data.

[0021] In addition to this, a computer-readable recording medium may be further provided for recording a computer program for executing a method for implementing the disclosed embodiments. Effects of the invention

[0023] According to the disclosed embodiments, a realistic and detailed pedestrian environment can be analyzed by constructing a map capable of precisely identifying various movement patterns, behavioral characteristics, interactions, etc., of pedestrians.

[0024] In addition, according to the disclosed embodiments, accurate and rapid reflection of real-time changes in the walking environment and pedestrian behavior can be implemented in a 3D precision walking map.

[0025] In addition, according to the disclosed embodiments, the efficiency and accuracy of constructing precision maps using public data can be improved, and the cost of spatial measurement for map creation and the map production period can be shortened. Brief explanation of the drawing

[0027] FIG. 1 is a block diagram illustrating a 3D precision pedestrian map production system according to one embodiment. FIGS. 2 to 6 are exemplary diagrams for explaining the process of producing a 3D precision pedestrian map according to one embodiment. FIG. 7 is a flowchart illustrating a method for producing a 3D precision pedestrian map according to one embodiment. FIG. 8 is a block diagram illustrating a computing environment including a computing device according to one embodiment. Specific details for implementing the invention

[0028] Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. The following detailed description is provided to facilitate a comprehensive understanding of the methods, apparatuses, and / or systems described herein. However, this is merely illustrative and the present invention is not limited thereto.

[0029] In describing the embodiments of the present invention, detailed descriptions of known technologies related to the present invention are omitted if it is determined that such detailed descriptions may unnecessarily obscure the essence of the present invention. Furthermore, the terms described below are defined in consideration of their functions within the present invention, and these may vary depending on the intentions or practices of the user or operator. Therefore, such definitions should be based on the content throughout this specification. Terms used in the detailed description are intended merely to describe the embodiments of the present invention and should not be limiting in any way. Unless explicitly stated otherwise, expressions in the singular form include the meaning of the plural form. In this description, expressions such as "include" or "comprise" are intended to refer to certain characteristics, numbers, steps, actions, elements, parts thereof, or combinations thereof, and should not be interpreted to exclude the existence or possibility of one or more other characteristics, numbers, steps, actions, elements, parts thereof, or combinations thereof other than those described.

[0030] FIG. 1 is a block diagram illustrating a 3D precision walking map production system according to one embodiment.

[0031] Hereinafter, the process of creating a 3D precision walking map according to one embodiment will be explained with reference to FIGS. 2 to 6, which are exemplary drawings.

[0032] The 3D precision walking map production system (100) can produce a 3D precision walking map in the order of generating a 3D terrain, objectifying the terrain, and modeling and placing additional objects (e.g., static walking obstacles).

[0033] Specifically, the 3D precision pedestrian map production system (100) can produce a 3D precision pedestrian map according to a step-by-step workflow of public data processing, processed data modeling, and modeling data transmission, and a detailed description thereof will be provided later. The above public data refers to various data collected during the operation of projects or research by government or public institutions, and the public data of the disclosed embodiment may refer to raw data provided for the construction of precision pedestrian maps, such as integrated map search and precision road maps provided by the National Land Information Platform. The public data is not limited to the data provided by the National Land Information Platform and may include all information collected for the production of precision pedestrian maps as needed by the operator.

[0034] Referring to FIG. 1, the 3D precision pedestrian map production system (100) includes a data processing unit (110), a data modeling unit (120), and a data providing unit (130). The components illustrated in FIG. 1 are not essential for implementing the 3D precision pedestrian map production system (100) according to the present disclosure, so the 3D precision pedestrian map production system (100) described in this specification may have more or fewer components than the components listed above.

[0035] The components illustrated in FIG. 1 may be connected to each other so as to be communicable through a communication network (not shown). In some embodiments, the communication network may include the Internet, one or more local area networks, wire area networks, cellular networks, mobile networks, other types of networks, or a combination of these networks.

[0036] The data processing unit (110) can obtain processed precision pedestrian map data by performing processing, including height application and normalization processing, on public data according to pre-set processing standards. That is, the data processing unit (110) can perform basic data processing using public data (used data in Table 1) to produce a 3D precision pedestrian map, and through the preliminary work, can extract processed precision pedestrian map data (generated data in Table 1) as shown in Table 1.

[0038] Usage data Generated data Processing details Boundary ContourLineOpen -Convert to EM format-Generate contour lines : Create ContourLineOpen(5186) RoadBasic Processed road boundary polygon -RoadBasic<Clip : Boundary> SidewalkBasic Processed sidewalk polygon -SidewalkBasic<Clip : Boundary> Water space river boundary Processed river boundary polygon - River boundary<Clip : Boundary> stream Processed Real Width River Polygon -Silpok Stream<Clip : Boundary> BoundaryRoad Processed plot polygon -BoundaryRoad<Clip : Boundary> : Create Boundary Plot - BoundaryPlot<Buffer : -0.5m> - Subtract: Fabricated road boundary, Fabricated sidewalk, Fabricated water space - Multiple parts into a single part - Field Calculator: Add Plot_ID Building Processing building -Modify Shapes-Delete Duplicate Shapes-Dissolve: Keep all properties except UFID-Building<Clip : Boundary> Create TargetBuilding - Multiple parts into a single part - TargetBuilding<Buffer : -0.1, 평평하게, miter> - <Attribute Table : $area is NULL> : Extract and delete null Geometry

[0039] The data processing unit (110) can predefine the 3D terrain of the 3D precise pedestrian map, including basic pedestrian environment objects such as roads, sidewalks, crosswalks, overpasses, parcels, traffic islands and water spaces (e.g., rivers, seas, etc.).

[0040] Referring to FIG. 2, the data processing unit (110) can obtain processed precision pedestrian map data through processing (200) including processing of public data related to terrain data (210), processing of building data (230), and processing of pedestrian obstacle data (230).

[0041] In the disclosed embodiments, walking obstacles may include static walking obstacles and dynamic walking obstacles, but temporarily existing dynamic walking obstacles may not be considered during the construction phase of the 3D precise walking map. Dynamic walking obstacles may be applied during the map utilization phase, such as walking simulations based on the constructed 3D precise walking map.

[0042] For example, static pedestrian obstruction elements may include bollards, street trees, lighting (streetlights / CCTV), flowerbeds, bus stops, subway stops, subway ventilation shafts, buildings, etc. Dynamic pedestrian obstruction elements may include pedestrians and vehicles, etc.

[0043] Referring to FIG. 2, the data modeling unit (120) can obtain modeling data by performing modeling processing (300) on processed precision walking map data according to preset modeling criteria including terrain group modeling (310), building modeling and placement (330), and walking obstacle modeling and placement (350).

[0044] The data modeling unit (120) can obtain an optimized terrain (TerrainOpt) using basic terrain data among processed precision walking map data when modeling a terrain group, obtain water space data through water space data modeling processing, and obtain a terrain group object (319 in FIG. 3) which is modeling data by processing a pre-set detail element modification modeling using at least one of the optimized terrain, water space data and additional terrain data.

[0045] Specifically, referring to FIG. 3, first, the data modeling unit (120) can receive basic terrain data (311) including contour line data (ContourLineOpen), processed road boundaries, processed sidewalks, processed river boundaries, processed rivers and processed parcels, and additional terrain data including processed crosswalks, processed traffic islands and other intermittent elements (processed stairs, processed overpasses, etc.).

[0046] Next, the data modeling unit (120) can generate a TerrainOpt (optimized terrain) using patch terrain generation.

[0047] For example, the data modeling unit (120) can generate patch terrain using the Patch (span 20) function with contour line data (ContourLineOpen), generate the necessary outermost map bounding rectangle using the contour line data, and generate optimized terrain (TerrainOpt) using the patch terrain data. At this time, the patch terrain data is in the form where only the bounding rectangle is extracted (<Clip : Bounding Rectangle> ) It may be. The data modeling unit (120) can convert the processed contour line data (ContourLineOpen) into an optimized 3D terrain (TerrainOpt) through the process described above.

[0048] Next, the data modeling unit (120) can create a waterfrontwall by performing the following water space data modeling (315) using the processed river boundary and the processed actual width river. If water space data does not exist, the water space data modeling can be omitted.

[0049] -<Trim with Region> Extract curves (ContourLineOpen Crv) from open contour data inside the processed river boundary using the function

[0050] A typical contour line is formed in a loop shape where the starting point and the ending point are the same, and in such cases, it can be described as closed. Based on this concept, the curve of the open contour data of the disclosed embodiment may mean that the loop is broken and the curve is open because the necessary area is cut and used as a boundary.

[0051] - Compare the z-values ​​of the endpoints of the extracted curves to extract the curve with the smallest z-value, and define that z-value as WaterMin Z.

[0052] - Create Points at 2m intervals along the processed river boundary curve and set them as WaterPoints.

[0053] - Project WaterPoint onto optimized 3D terrain (TerrainOpt)

[0054] - Extract the point with the highest z-value among the projected points and define that z-value as WaterMaxZ.

[0055] - WaterDepth extraction satisfying

[0056] - After projecting the subsequently modeled waterfront curve onto TerrainOpt<Extrude : -WaterDepth> Create a waterfront wall through the work

[0057] The above-mentioned waterside space may also be referred to as a waterside. The above-mentioned waterside space may refer to a space near water, such as Han River Park or the Cheonggyecheon walking path. In the disclosed embodiment, the waterside space may refer to terrain near water spaces, such as rivers, that constitute a city.

[0058] The above waterfront slope may refer to a slope that fills the gap in elevation between the waterfront space (Waterfront) and the terrain on the bridge.

[0059] Next, the data modeling unit (120) can create a terrain group object (319) by performing a modeling process for modifying the detail elements (317) of the data generated by applying an automation algorithm. The modeling process for modifying the detail elements (317) may be as follows.

[0060] - Road modeling

[0061] : Process TerrainOpt, split according to road boundary geometry, and extract road portions

[0062] : Split the road section into a processed sidewalk once again

[0063] : Generate a road with optimized capacity by converting the result to Nurbs to Mesh (function-based modeling → 3D Mesh Geometry) (road geometry)

[0064] The above road geometry is a concept that defines data types and can be defined as road modeling.

[0065] - Press Modeling

[0066] : Split TerrainOpt into processed sidewalks to extract the sidewalk portion

[0067] :<Extrude : 0.15> After inserting depth values ​​into the cut surface, generate sidewalk geometry through Nurbs to Mesh operation

[0068] - Parcel Modeling

[0069] : Split TerrainOpt into processed parcels and extract parcel portions

[0070] :<Extrude : 0.15> Subsequently, plot generation (plot geometry) through Nurbs to Mesh operation

[0071] - Water space modeling (skip if absent)

[0072] : Process TerrainOpt and split it at the river boundary to create water terrain

[0073] : Water terrain in the Z direction<Move Srf : -WaterDepth> Execute task

[0074] : Process WaterTerrain and split it again into a real-width stream to create Water and waterfront geometry.

[0075] : Water :<Extrude : -0.3, not solid> After the work<Move Srf : -0.3>

[0076] : Both Water and Waterfront geometries are transformed via Nurbs to Mesh operation

[0077] - Traffic island modeling (skip if unavailable)

[0078] Split road data into traffic islands using the Meshsplit command.

[0079] :<Extrude : 0.15> Subsequently, Island Geometry is generated through the Nurbs to Mesh operation.

[0080] - Crosswalk Modeling

[0081] : Modified the model by processing individual crosswalk data to ensure that a surface inevitably meets the road boundary.

[0082] Stripe Pattern Modeling

[0083] :<Meshsplit : Road, 가공 횡단보도> Create Crossroad Geometry through the process

[0084] - Overpass modeling (skip if absent)

[0085] : Rhino Modeling Based on Processed Overpass Data

[0086] : Generate Overpass Geometry through Nurbs to Mesh operation

[0087] - Stair modeling (skip if absent)

[0088] :<Meshsplit : Road, 가공 계단> Create Stairs data through the task

[0089] Stairs Geometry generated via Nurbs to Mesh operation after Rhino Modeling

[0090] - Road separation

[0091] Divide the entire Road Object into Road_Both (side roads), Road_Car (roads or main roads where pedestrians cannot walk), and Road_Out (areas outside the interaction range).

[0092] : Created so that when generating Road_Both, parts where the pedestrian network connects without crosswalks are included.

[0094] Below, we will explain using building modeling and layout as examples.

[0095] The data modeling unit (120) can perform modeling and terrain placement operations using processed building data generated during the public data processing stage when modeling and placing buildings. As shown in FIG. 4, the data modeling unit (120) can create and execute a building arrangement algorithm.

[0096] Referring to FIG. 4, the data modeling unit (120) can extract a Z value (Z Value) (332) according to a pre-set first modeling standard using processed building data (331) among processed precision pedestrian map data when modeling and arranging a building. Specifically, the data modeling unit (120) can perform the following procedure.

[0097] - Import processed building data

[0098] - After extracting boundary (line) data of the processed building polygon, extract vertices using the Deconstruct Brep command.

[0099] - Create points by projecting vertices onto the Path Terrain using the Project command

[0100] - Extract Z Value using the Deconstruct Point command

[0101] The above Z Value extraction may mean extracting the Z value of the Geometry, that is, the elevation value.

[0102] The data modeling unit (120) can classify a list of Z value data containing at least one extracted Z value to extract a minimum Z value and a maximum Z value, and use this to define the landscape height (333).

[0103] The data modeling unit (120) can generate a Building Basic Height (334) by applying a preset average height per floor to the level basic attribute of the processed building data.

[0104] For example, the data modeling unit (120) can obtain the basic building height by multiplying the basic level attribute of the processed building data by a value assuming a preset average height per floor of 4m.

[0105] The data modeling unit (120) can generate a building height (335) using the landscape elevation height and the building base height.

[0106] Specifically, the data modeling unit (120) can extract a Building Height value that satisfies a mathematical formula such as BuildingHeight = LandscapeHeight + BuildingBasicHeight, when the final height assuming the building is placed in the Z direction on the terrain is defined as BuildingHeight.

[0107] The data modeling unit (120) can extrude the building projection plane (BuildingSrf(Building Surface)) data to the height of the building, and move and arrange the extruded data according to a preset second modeling standard to complete the building arrangement (336). The building projection plane may also be named building GIS polygon data.

[0108] Specifically, building data in public data can be represented as polygon data (2D data) in the form of a projected building from above onto the floor. Extruding the building projection plane (BuildingSrf) data by the height may mean creating a 3D form (raising it three-dimensionally) of the flat building polygon data by the height of each building. Afterwards, the data modeling unit (120) can perform the process of tracking and arranging the coordinate positions.

[0110] For example, the data modeling unit (120) extrudes data<MoveZ : MinZ + 0.15> You can complete the execution of the BuildingArrangement algorithm by moving and arranging through the task.

[0112] Below, we will explain the modeling and placement of walking obstacle elements using an example.

[0113] The data processing unit (110) can obtain relatively equal public data through previous procedures and select four objects as pedestrian environment elements that are easily accessible in a pedestrian environment, such as traffic lights, bollards, trees, and lighting (streetlights, CCTV).

[0114] As shown in Table 2, the data processing unit (110) can solve the problem of improving accuracy and optimizing pedestrian obstacle elements by processing public data first. The data processing unit (110) can obtain processed precision pedestrian map data (generated data in Table 2) by processing public data of pedestrian obstacle elements (used data in Table 2).

[0116] Usage data Generated data Processing details bollard Obstacle bollard - Identify bollard locations in the target area using a map platform - Generate obstacle bollard data - Build data centered on intersections and main roads blinker Obstacle_Traffic_Light -SignalBasic Layer Processing: Obstacle_Traffic Light Data Generation - Verification of Data Position Errors and Adjustment of Deviated Data Positions colonnade Obstacle_Street Trees -Edit Shape-Delete Duplicate Shape-Street Tree<Clip : Boundary> Obstacle_StreetTree Data Generation - Layer Processing lighting Obstacle_Lighting -Edit Shapes-Delete Duplicate Shapes-Lighting<Clip : Boundary> -Delete duplicates with parcel data-Delete Null Geometry-Field Calculator: Add Light_ID-Layer Processing: Generate Obstacle_Lighting Data

[0117] The data modeling unit (120) can apply an automated placement algorithm to place the processed precision walking map data (generated data in Table 2) obtained through processing the above-described public walking obstacle data (used data in Table 2) on the terrain after performing individual modeling and rendering work.

[0118] Referring to FIG. 5, the data modeling unit (120) can perform processing (355) to place pedestrian obstacle data (351) (e.g., Table 2) among processed precision pedestrian map data on the terrain according to a preset automated placement algorithm when modeling and placing pedestrian obstacles. This includes applying a FlipMatrix texture mapping (352), aligning the terrain Z coordinates based on MeshRay (353, 354), and placing the data so that it is perpendicular to the direction of travel of the road. Through this, the data modeling unit (120) can perform the placement of pedestrian obstacles (356).

[0119] The data modeling unit (120) can place walking obstacle data on the terrain according to the following automated placement algorithm.

[0120] - Automated batch algorithm

[0121] import obstacle_bollard, obstacle_traffic_light, obstacle_street_tree

[0122] : Performed X, Y coordinate alignment with the testbed precision walking map constructed using transfer data

[0123] Assuming elements such as Sidewalk and Island are the Target Area, perform Z-coordinate alignment of the data using the MeshRay command.

[0124] : Apply mapping data using the MaterialAttributes function, and utilize the FlipMatrix function to enable one-to-one mapping for each object's individual constituent parts during future mapping operations.

[0125] If necessary, create additional modeling parts to manage gaps or unevenness that may occur between objects and walking terrain.

[0126] For roadside tree objects, use the Orient function to position them perpendicular to the direction of travel.

[0128] After terrain generation, building generation and placement, and pedestrian obstacle generation and placement are completed by the data modeling unit (120), an example of a rendering image after setting a Material through texture mapping according to the components of each object may be as shown in FIG. 6. At this time, FIG. 6 may be an example of a rendering image of a 3D precise pedestrian map of the Seorae Village area.

[0129] The data providing unit (130) can perform a procedure to transmit modeling data to an engine program capable of data expansion and system application so that the modeling data obtained by processing precision walking map data according to pre-set modeling criteria can be utilized for various purposes (e.g., use as a digital twin space, path search, application to a walking simulation system, etc.). At this time, the data providing unit (130) can perform a data conversion operation to transmit the modeling data to the engine program.

[0130] The data providing unit (130) provides modeling data, and can generate and provide a 3D precision walking map by converting the modeling data according to a preset data conversion standard.

[0131] The data providing unit (130) can convert the modeling data according to a data conversion standard that includes at least one of spatial error optimization and 3D coordinate range optimization.

[0132] Specifically, the data provider (130) can ensure that no positional relationship error occurs when objects are aggregated into one group during spatial error optimization, and can construct a 3D precise pedestrian map with the coordinate system of the public data unified.

[0133] When optimizing the 3D coordinate range, the data providing unit (130) can replace the latitude and longitude coordinates of the coordinate system with the X and Y values ​​of the 3D space by comparing them with the origin (0, 0, 0) of the modeling program.

[0134] The data provider (130) can apply transfer data collectively by calculating the data based on moving the entire object to the origin, in order to prevent problems such as being unrecognizable or having an excessive capacity in the engine program, considering that the size and unit of the entire map may become relatively excessively large depending on the coordinate system when optimizing the 3D coordinate range.

[0135] The 3D precision pedestrian map creation service may be provided, for example, in the form of a web page or through an application installed on a user device (not shown).

[0136] According to one embodiment, the 3D precision walking map production system (100) may be implemented as a separate configuration capable of communicating with a server that provides various services based on the 3D precision walking map, but is not limited thereto, and may be integrated and implemented as a configuration within a server that provides various services based on the 3D precision walking map.

[0137] According to one embodiment, the data processing unit (110), the data modeling unit (120), and the data providing unit (130) may each be implemented using one or more physically separated devices, or by one or more hardware processors or a combination of one or more hardware processors and software, and unlike the illustrated example, they may not be clearly distinguished in specific operation.

[0138] The 3D precision walking map generation system (100) of the present disclosure may be composed of one or more cores and may include a processor (not shown) for data analysis and deep learning, such as a central processing unit of a computing device, a general purpose graphics processing unit, or a tensor processing unit. The 3D precision walking map generation system (100) may read a computer program stored in memory (not shown) and perform data processing for machine learning according to the present disclosure. According to the present disclosure, the processor may perform calculations for learning a neural network. The processor may perform calculations for learning a neural network, such as processing input data for learning in deep learning, extracting features from input data, calculating errors, and updating the weights of the neural network using backpropagation.

[0139] The above neural network model may be a deep neural network. In this disclosure, neural network, network function, and neural network may be used interchangeably. A deep neural network (DNN) may refer to a neural network that includes multiple hidden layers in addition to an input layer and an output layer. Using a deep neural network allows for the identification of latent structures of data. That is, it is possible to identify the latent structures of photos, text, videos, voice, and music (e.g., what objects are in a photo, what the content and emotions of the text are, what the content and emotions of the voice are, etc.). A deep neural network may include a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a Q network, a U network, a Siamese network, etc.

[0140] Convolutional Neural Networks (CNNs) are a type of deep neural network that includes neural networks containing convolutional layers. A CNN is a type of multilayer perceptron designed to use minimal preprocessing. A CNN can consist of one or more convolutional layers and artificial neural network layers combined with them. CNNs can additionally utilize weights and pooling layers. Thanks to this structure, CNNs can fully utilize input data with a two-dimensional structure. Convolutional neural networks can be used to recognize objects in images. Convolutional neural networks can process image data by representing it as a matrix with dimensions. For example, in the case of RGB (red-green-blue) encoded image data, it can be represented as a two-dimensional matrix for each of the R, G, and B colors (e.g., in the case of a two-dimensional image). That is, the color value of each pixel in the image data can be an element of the matrix, and the size of the matrix can be equal to the size of the image. Therefore, image data can be represented as three 2-dimensional matrices (a 3-dimensional data array).

[0141] In a convolutional neural network, the convolutional process (input / output of the convolutional layer) can be performed by moving the convolutional filter and multiplying the matrix components of the convolutional filter at each location in the image. The convolutional filter can be composed of an n*n matrix. Generally, the convolutional filter can be composed of a fixed-size filter smaller than the total number of pixels in the image. That is, when an m*m image is input into a convolutional layer (e.g., a convolutional layer with an n*n size convolutional filter), a matrix representing n*n pixels containing each pixel of the image can be multiplied by the convolutional filter and its components (i.e., multiplying the matrix components). Through this multiplication with the convolutional filter, components matching the convolutional filter can be extracted from the image. For example, a 3x3 convolutional filter for extracting vertical linear components from an image can be constructed as [[0,1,0], [0,1,0], [0,1,0]]. When a 3x3 convolutional filter for extracting vertical linear components is applied to an input image, vertical linear components matching the convolutional filter can be extracted and output from the image. A convolutional layer can apply a convolutional filter to each matrix for each channel representing the image (i.e., R, G, B colors in the case of an R, G, B coded image). A convolutional layer can extract features matching the convolutional filter from the input image by applying the convolutional filter to the input image. The filter values ​​of the convolutional filter (i.e., the values ​​of each component of the matrix) can be updated by backpropagation during the training process of the convolutional neural network.

[0142] A subsampling layer may be connected to the output of a convolutional layer to simplify the output of the convolutional layer, thereby reducing memory usage and computational load. For example, when the output of a convolutional layer is input to a pooling layer having a 2x2 max pooling filter, the image can be compressed by outputting the maximum value contained in each patch for every 2x2 patch from each pixel of the image. The aforementioned pooling may be a method of outputting a minimum value from a patch or an average value from a patch, and any pooling method may be included in the present disclosure.

[0143] A convolutional neural network may include one or more convolutional layers and subsampling layers. A convolutional neural network can extract features from an image by repeatedly performing convolutional processes and subsampling processes (e.g., the aforementioned max pooling). Through repeated convolutional and subsampling processes, the neural network can extract global features of the image.

[0144] The output of a convolutional layer or a subsampling layer can be input to a fully connected layer. A fully connected layer is a layer where every neuron in one layer is connected to every neuron in an adjacent layer. In a neural network, a fully connected layer can refer to a structure where every node in each layer is connected to every node in another layer.

[0145] At least one of the CPU, GPGPU, and TPU of the processor can process the training of a network function. For example, the CPU and GPGPU can together process the training of a network function and data classification using the network function. In addition, in one embodiment of the present disclosure, processors of a plurality of computing devices can be used together to process the training of a network function and data classification using the network function. Furthermore, a computer program executed in a computing device according to one embodiment of the present disclosure may be a CPU, GPGPU, or TPU executable program.

[0147] FIG. 7 is a flowchart for explaining a method for creating a 3D precision walking map according to one embodiment. The method illustrated in FIG. 7 can be performed, for example, by the aforementioned 3D precision walking map creation system (100). Although the method is described in the illustrated flowchart by dividing it into a plurality of steps, at least some of the steps may be performed in a different order, combined with other steps and performed together, omitted, divided into detailed steps, or performed with one or more steps not illustrated added.

[0148] The 3D precision walking map production method described below can apply all operations of the 3D precision walking map production system (100) disclosed by referring to FIGS. 1 to 5 described above, but for the convenience of explanation, redundant explanations will be omitted.

[0149] In step 1100, the 3D precision pedestrian map production system (100) can obtain processed precision pedestrian map data by performing processing, including height application and normalization processing, on public data through the data processing unit (110) according to preset processing standards.

[0150] The 3D precision pedestrian map production system (100) can obtain processed precision pedestrian map data through a data processing unit (110) by processing public data including terrain-related data processing, building data processing, and pedestrian obstacle data processing.

[0151] In step 1200, the 3D precision pedestrian map production system (100) can obtain modeling data by processing the processed precision pedestrian map data through the data modeling unit (120) according to pre-set modeling criteria including terrain group modeling, building modeling and placement, and pedestrian obstacle element modeling and placement.

[0152] In step 1300, the 3D precision pedestrian map production system (100) provides modeling data through the data provision unit (130), and can generate and provide a 3D precision pedestrian map by converting the modeling data according to a preset data conversion standard.

[0153] Meanwhile, although not yet done, prior to step 1100, the 3D precision pedestrian map production system (100) can predefine the 3D terrain of the 3D precision pedestrian map, including basic pedestrian environment objects such as roads, sidewalks, crosswalks, overpasses, parcels, traffic islands, and water spaces, through the data processing unit (110).

[0155] FIG. 8 is a block diagram illustrating a computing environment including a computing device according to one embodiment. In the illustrated embodiment, each component may have different functions and capabilities in addition to those described below, and may include additional components in addition to those described below.

[0156] The illustrated computing environment (10) includes a computing device (12). The computing device (12) may be one or more components included in a 3D precision walking map creation system (100) according to one embodiment.

[0157] The computing device (12) includes at least one processor (14), a computer-readable storage medium (16), and a communication bus (18). The processor (14) can cause the computing device (12) to operate according to the exemplary embodiment described above. For example, the processor (14) can execute one or more programs stored in the computer-readable storage medium (16). The one or more programs may include one or more computer-executable instructions, and the computer-executable instructions may be configured to cause the computing device (12) to perform operations according to the exemplary embodiment when executed by the processor (14).

[0158] A computer-readable storage medium (16) is configured to store computer-executable instructions or program code, program data and / or other suitable forms of information. A program (20) stored in the computer-readable storage medium (16) includes a set of instructions executable by a processor (14). In one embodiment, the computer-readable storage medium (16) may be memory (volatile memory such as random access memory, non-volatile memory, or a suitable combination thereof), one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or other forms of storage media that are accessed by a computing device (12) and capable of storing desired information, or a suitable combination thereof.

[0159] The communication bus (18) interconnects various other components of the computing device (12), including the processor (14) and the computer-readable storage medium (16).

[0160] The computing device (12) may also include one or more input / output interfaces (22) and one or more network communication interfaces (26) that provide interfaces for one or more input / output devices (24). The input / output interfaces (22) and network communication interfaces (26) are connected to a communication bus (18). The input / output devices (24) may be connected to other components of the computing device (12) through the input / output interfaces (22). An exemplary input / output device (24) may include an input device such as a pointing device (such as a mouse or trackpad), a keyboard, a touch input device (such as a touchpad or touchscreen), a voice or sound input device, various types of sensor devices and / or imaging devices, and / or an output device such as a display device, a printer, a speaker and / or a network card. An exemplary input / output device (24) may be included inside the computing device (12) as a component constituting the computing device (12), or it may be connected to the computing device (12) as a separate device distinct from the computing device (12).

[0161] The disclosed embodiments may be implemented in the form of a recording medium that stores instructions executable by a computer. The instructions may be stored in the form of program code and, when executed by a processor, may generate a program module to perform the operation of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium.

[0162] Although representative embodiments of the present invention have been described in detail above, those skilled in the art will understand that various modifications can be made to the above-described embodiments without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims set forth below as well as equivalents thereof. Explanation of the symbols

[0163] 10: Computing Environment 12: Computing device 14: Processor 16: Computer-readable storage media 18: Communication bus 20: Program 22: Input / Output Interface 24: Input / Output Devices 26: Network communication interface 100: 3D Precision Pedestrian Mapping System 110: Data Processing Unit 120: Data Modeling Department 130: Data Provider

Claims

Claim 1 A 3D precise pedestrian map production system comprising: a data processing unit that obtains processed precise pedestrian map data by performing processing on public data, including height application and normalization processing, according to preset processing standards; a data modeling unit that obtains modeling data by performing modeling processing on the processed precise pedestrian map data according to preset modeling standards, including terrain group modeling, building modeling and placement, and pedestrian obstacle modeling and placement; and a data providing unit that provides the modeling data, and generates and provides a 3D precise pedestrian map by converting the modeling data according to preset data conversion standards. Claim 2 A 3D precise pedestrian map production system according to claim 1, wherein the data processing unit predefines the 3D terrain of the 3D precise pedestrian map, including basic pedestrian environment objects such as roads, sidewalks, crosswalks, overpasses, parcels, traffic islands, and water spaces. Claim 3 A 3D precision pedestrian map production system according to claim 1, wherein the data processing unit obtains the processed precision pedestrian map data through the processing of the public data, including terrain-related data processing, building data processing, and pedestrian obstacle data processing. Claim 4 A 3D precision walking map production system according to claim 1, wherein the data modeling unit, when modeling the terrain group, obtains an optimized terrain (TerrainOpt) using basic terrain data among the processed precision walking map data, obtains water space data through water space data modeling processing, and obtains a terrain group object which is the modeling data by processing a pre-set detail element modification modeling using at least one of the optimized terrain, the water space data, and additional terrain data. Claim 5 A 3D precision pedestrian map production system according to claim 1, wherein the data modeling unit, when performing building modeling and placement, extracts a Z value according to a preset first modeling standard using processed building data among the processed precision pedestrian map data, classifies a list of Z value data including at least one extracted Z value to extract a minimum Z value and a maximum Z value, defines a landscape height using the same, generates a building basic height by applying a preset average height per floor to the basic attribute of the number of floors among the processed building data, generates a building height using the landscape height and the building basic height, extrudes BuildingSrf data by the building height, and completes a building arrangement by moving and placing the extruded data according to a preset second modeling standard. Claim 6 A 3D precision pedestrian map production system according to claim 1, wherein the data modeling unit, when modeling and placing pedestrian obstacle elements, places pedestrian obstacle element data among the processed precision pedestrian map data onto the terrain according to a preset automated placement algorithm, and performs processing such as applying FlipMatrix texture mapping, aligning terrain Z coordinates based on MeshRay, and placing the data so as to be perpendicular to the road direction of travel. Claim 7 A 3D precision walking map production system according to claim 1, wherein the data providing unit converts the modeling data according to the data conversion standard including at least one of spatial error optimization and 3D coordinate range optimization. Claim 8 A method for producing a 3D precise pedestrian map, comprising: a step of obtaining processed precise pedestrian map data by performing processing on public data, including height application and normalization processing according to preset processing standards; a step of obtaining modeling data by performing modeling processing on the processed precise pedestrian map data according to preset modeling standards, including terrain group modeling, building modeling and placement, and pedestrian obstacle modeling and placement; and a step of providing the modeling data, wherein the modeling data is converted according to preset data conversion standards to generate and provide a 3D precise pedestrian map. Claim 9 A method for producing a 3D precise pedestrian map according to claim 8, further comprising the step of pre-defining the 3D terrain of the 3D precise pedestrian map, including basic pedestrian environment objects such as roads, sidewalks, crosswalks, overpasses, parcels, traffic islands, and water spaces. Claim 10 A method for producing a 3D precision pedestrian map according to claim 8, wherein, in the step of acquiring the processed precision pedestrian map data, the processed precision pedestrian map data is acquired through the processing of the public data including terrain-related data processing, building data processing, and pedestrian obstacle data processing.

Images