Motor vehicle parking lot scene map construction method and device and computer readable storage medium

Abstract

Embodiments of the present application provide a motor vehicle parking lot scene map construction method, device and computer readable storage medium, the method comprising: constructing a global map coordinate system with a preset starting point of the parking lot as the coordinate origin, controlling the motor vehicle to drive from the preset starting point to the preset terminal point; in the process of driving the motor vehicle, generating a panoramic overhead view based on the video images collected by each vehicle-mounted camera; obtaining the current position coordinates of the motor vehicle at a predetermined period and constructing a local scene map of the motor vehicle based on the physical coordinates of the predetermined target in the panoramic overhead view; mapping the local scene map constructed in each period in the global map coordinate system according to the current position coordinates obtained in the corresponding period to obtain an initial global scene map; and excluding the repeated physical coordinates of the same predetermined target in the initial global scene map to obtain a target global scene map. The scene map constructed in this embodiment can be used to assist in selecting a parking space and to assist in realizing path planning.

Description

TECHNICAL FIELD

[0001] The embodiment of the application relates to the technical field of motor vehicle mapping, in particular to a motor vehicle parking lot scene map construction method, device and computer readable storage medium. BACKGROUND

[0002] With the gradual improvement of the intelligence degree of motor vehicles, motor vehicles begin to be configured with automatic parking systems, in order to realize accurate path planning and parking navigation, the existing automatic parking systems usually need to construct a parking lot scene map of a motor vehicle from a preset starting point to a preset terminal point of a parking lot.

[0003] A construction method of a parking lot scene map for an automatic parking system mainly adopts image data photographed by a vehicle-mounted camera and a visual SLAM technology to realize the construction, in the process of map construction, semantic information of specific targets such as parking spaces, lane lines and pillars in the parking lot needs to be extracted, and the semantic information is fused into the constructed parking lot scene map.

[0004] However, the inventors have found in the specific implementation that after the semantic information of the specific targets is extracted, the semantic information is not further processed and is directly fused into the parking lot scene map, so that the related semantic information is difficult to accurately reflect the exact positions of the specific targets, the constructed parking lot scene map cannot be used to guide the selection of parking spaces and realize path planning, and finally the constructed parking lot scene map can only realize the automatic parking task of a fixed track and cannot flexibly realize the automatic parking of a dynamic scene according to the change of a scene. SUMMARY

[0005] The technical problem to be solved by the embodiment of the application is to provide a motor vehicle parking lot scene map construction method, and the constructed scene map can be used to assist the selection of parking spaces and assist the realization of path planning.

[0006] The technical problem to be further solved by the embodiment of the application is to provide a motor vehicle parking lot scene map construction device, and the constructed scene map can be used to assist the selection of parking spaces and assist the realization of path planning.

[0007] The technical problem to be further solved by the embodiment of the application is to provide a computer readable storage medium, which can store a computer program of the constructed scene map, and the computer program can be used to assist the selection of parking spaces and assist the realization of path planning.

[0008] In order to solve the above technical problems, the embodiment of the application provides the following technical scheme: a motor vehicle parking lot scene map construction method, comprising the following steps:

[0009] construct a global map coordinate system with a preset starting point of the parking lot as an original point, and control the motor vehicle to drive from the preset starting point to a preset ending point along a preset driving path;

[0010] During the driving of the motor vehicle, control each vehicle-mounted camera of the motor vehicle to collect video images of the surrounding environment of the motor vehicle in real time, and generate a panoramic overhead view based on the video images collected by each vehicle-mounted camera;

[0011] acquire a current position coordinate of the motor vehicle in the global map coordinate system at a predetermined period and construct a local scene map based on the current position coordinate, the construction process of the local scene map comprising: extracting semantic information of a predetermined target in the parking lot from the panoramic overhead view, processing the semantic information to obtain physical coordinates of the predetermined target having an actual physical length, and constructing a local scene map of the motor vehicle at the current position coordinate based on the physical coordinates in the panoramic overhead view, the predetermined target at least including parking spaces, lane lines, and pillars;

[0012] map the local scene map constructed in each period in the global map coordinate system according to the current position coordinate obtained in the corresponding period to obtain an initial global scene map; and

[0013] exclude the physical coordinates of the same predetermined target that are repeated in the initial global scene map to obtain a target global scene map of the parking lot.

[0014] Further, the semantic information of the predetermined target refers to pixel regions of parking space lines, pixel regions of lane lines, and pixel regions of pillars obtained by performing semantic segmentation on the panoramic overhead view using a predetermined deep network model, and the processing of the semantic information to obtain physical coordinates of the predetermined target having an actual physical length specifically comprises:

[0015] performing morphological operations on the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars respectively to fill pixel holes in the pixel regions;

[0016] performing image processing on the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars after the morphological operations to obtain pixel coordinates of each parking space, pixel coordinates of each lane line, and pixel coordinates of each pillar in the panoramic overhead view, respectively; and

[0017] convert the pixel coordinates of each parking space, the pixel coordinates of each lane line, and the pixel coordinates of each pillar in the panoramic overhead view into corresponding physical coordinates, respectively.

[0018] Further, the pixel coordinates of the parking space are obtained by the following steps: sequentially adopting a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the pixel region of the parking space line after the morphological operation to obtain a plurality of parking space lines, and constructing the parking space and its pixel coordinates in the panoramic overhead view based on the relative position relationship of each of the parking space lines.

[0019] Further, the pixel coordinates of the lane line are obtained by the following steps: sequentially adopting a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the pixel region of the lane line after the morphological operation to obtain a plurality of suspected lines, and excluding interference lines with a length less than a preset length threshold and a horizontal distribution from each of the suspected lines to obtain each lane line and its pixel coordinates in the panoramic overhead view.

[0020] Further, the pixel coordinates of the column are obtained by the following steps: calculating the shortest pixel distance between each pixel point in the pixel region of the column and the parking space line and the lane line, removing the pixel points in the pixel region of the column with the shortest pixel distance greater than a preset pixel distance, and sequentially adopting a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the remaining pixel points in the pixel region of the column to obtain the connecting line of the column and the parking lot ground, and constructing the column and its pixel coordinates in the panoramic overhead view based on the connecting line.

[0021] Further, the exclusion of the physical coordinates of the same predetermined target in the initial global scene map specifically includes:

[0022] Excluding the remaining physical coordinates in the global scene map that belong to the same parking space and excluding the latest physical coordinates;

[0023] Excluding the remaining physical coordinates in the global scene map that belong to the same column and excluding the latest physical coordinates; and

[0024] Combining the physical coordinates in the global scene map that belong to the same lane line using a predetermined straight line clustering algorithm model.

[0025] Further, after excluding the physical coordinates of the same predetermined target in the initial global scene map, the initial global scene map is further image-constrained to obtain the target global scene map, and the image constraint includes the same size constraint and lane line alignment constraint.

[0026] Further, after constructing the local scene map of the motor vehicle at the current position coordinate based on the physical coordinates, the use state of the parking space in the local scene map is also marked, and the use state of the parking space is mapped into the target global scene map. Each time the target global scene map is used for auxiliary driving, the use state of the parking space in the target global scene map is also updated. The method for marking the use state of the parking space in the local scene map and updating the use state of the parking space in the target global scene map comprises:

[0027] obtaining the coordinate data of the obstacle vehicle in the global map coordinate system based on the predetermined vehicle recognition model;

[0028] obtaining the coordinate data of the obstacle vehicle in the global map coordinate system based on the predetermined vehicle recognition model;

[0029] marking or updating the use state of the parking space according to the coordinate data of the obstacle vehicle.

[0030] In another aspect, to solve the above technical problems, the embodiments of the present application provide the following technical solutions: a motor vehicle parking lot scene map construction device connected with a driving system of a motor vehicle and each vehicle-mounted camera, the device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the motor vehicle parking lot scene map construction method according to any one of the above when executing the computer program.

[0031] In another aspect, to solve the above technical problems, the embodiments of the present application provide the following technical solutions: a computer readable storage medium comprising a stored computer program, wherein the computer readable storage medium controls the device where the computer readable storage medium is located to execute the motor vehicle parking lot scene map construction method according to any one of the above when the computer program runs.

[0032] By adopting the technical scheme, the embodiment of the present application has at least the following beneficial effects: the embodiment of the present application extracts semantic information of predetermined targets such as parking spaces, lane lines and pillars in a parking lot from a panoramic overhead view, and then further processes the semantic information to obtain physical coordinates of the predetermined targets having actual physical lengths, and constructs a local scene map of the motor vehicle at a current position coordinate based on the physical coordinates, so that the parking spaces, lane lines and pillars in the local scene map constructed, and the initial global scene map and the target global scene map further generated based on the local scene map, all contain corresponding physical coordinates, thereby when the automatic parking system uses the target global scene map finally constructed, the physical coordinates of the parking spaces, lane lines and pillars can be used for dynamic selection of parking spaces and dynamic planning of parking paths, which is beneficial to improving the efficiency and success rate of automatic parking. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 A step flow chart of an optional embodiment of the motor vehicle parking lot scene map construction method of the present application.

[0034] Figure 2 A principle block diagram of an optional embodiment of the motor vehicle parking lot scene map construction device of the present application.

[0035] Figure 3 A functional module diagram of an optional embodiment of the motor vehicle parking lot scene map construction device of the present application. DETAILED DESCRIPTION

[0036] The present application will be further described in detail below in conjunction with the drawings and specific embodiments. It should be understood that the following illustrative embodiments and descriptions are only used to explain the present application and are not intended to limit the present application, and the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0037] As shown in Figure 1 An optional embodiment of the present application provides a motor vehicle parking lot scene map construction method, comprising the following steps:

[0038] S1: constructing a global map coordinate system with a preset starting point of a parking lot as a coordinate origin, and controlling a motor vehicle to travel from the preset starting point to a preset terminal point according to a preset travel path;

[0039] S2: during the travel of the motor vehicle, controlling each vehicle-mounted camera 1 of the motor vehicle to collect video images of the environment around the motor vehicle in real time, and generating a panoramic overhead view based on the video images collected by each vehicle-mounted camera;

[0040] S3: acquiring a current position coordinate of the motor vehicle in the global map coordinate system in a predetermined period and constructing a local scene map based on the current position coordinate, the construction process of the local scene map comprising: extracting semantic information of predetermined targets in the parking lot from the panoramic overhead view, processing the semantic information to obtain physical coordinates of the predetermined targets having actual physical lengths, and constructing a local scene map of the motor vehicle at the current position coordinate based on the physical coordinates in the panoramic overhead view, the predetermined targets at least comprising parking spaces, lane lines and pillars;

[0041] S4: mapping the local scene map constructed in each period in the global map coordinate system according to the current position coordinate obtained in the corresponding period to obtain an initial global scene map; and

[0042] S5: excluding the physical coordinates of the same predetermined target repeatedly in the initial global scene map to obtain a target global scene map of the parking lot.

[0043] The embodiment of the present application extracts semantic information of predetermined targets such as parking spaces, lane lines and pillars in the parking lot from the panoramic overhead view, and then further processes the semantic information to obtain physical coordinates of the predetermined targets having actual physical lengths, and constructs a local scene map of the motor vehicle at the current position coordinate based on the physical coordinates, so that the parking spaces, lane lines and pillars in the local scene map constructed, the initial global scene map and the target global scene map further generated based on the local scene map all contain corresponding physical coordinates, so that when the automatic parking system uses the target global scene map finally constructed, the physical coordinates of the parking spaces, lane lines and pillars can be used for dynamic selection of parking spaces and dynamic planning of parking paths, which is beneficial to improve the efficiency and success rate of automatic parking.

[0044] In the embodiment, in step S1, when the global map coordinate system is constructed, a preset starting point (the geometric center of the motor vehicle when the motor vehicle starts to run is also located at the preset starting point) is taken as the coordinate origin, the head direction is taken as the positive direction of the y axis, and the direction perpendicular to the y axis to the right is taken as the positive direction of the x axis. In addition, the motor vehicle can be driven manually or automatically controlled by an intelligent driving system. In step S2, when the panoramic overhead view is formed based on the video images collected by each vehicle-mounted camera 1, the camera extrinsic parameters of the vehicle-mounted camera 1 are used to convert each frame of image of the four video images into an overhead field of view through perspective transformation, and then the images of the overhead field of view are spliced and fused to form the panoramic overhead view. In step S3, when the local scene map is constructed, a two-dimensional coordinate system is constructed with the geometric center of the vehicle as the origin, the right direction as the positive direction of the x axis, and the upward direction as the positive direction of the y axis, and then the detected parking spaces, lane lines, and pillars are marked in the two-dimensional coordinate system according to the physical coordinates, so as to obtain the local scene map containing semantics. In step S4, in addition, the period can be every predetermined time length (for example, 5 s) when the motor vehicle runs at a constant speed or every driving distance (for example, 50 cm) of the motor vehicle, and the period parameter can be obtained from the odometer of the motor vehicle.

[0045] In an optional embodiment of the present application, the semantic information of the predetermined target refers to the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars obtained by performing semantic segmentation on the panoramic overhead view by using a predetermined deep network model, and the processing of the semantic information to obtain the physical coordinates with actual physical lengths of the predetermined targets specifically includes:

[0046] Performing morphological operations on the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars to fill pixel holes in the pixel regions;

[0047] Performing image processing on the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars after the morphological operations to obtain the pixel coordinates of each parking space, the pixel coordinates of each lane line, and the pixel coordinates of each pillar in the panoramic overhead view, respectively; and

[0048] Converting the pixel coordinates of each parking space, the pixel coordinates of each lane line, and the pixel coordinates of each pillar in the panoramic overhead view into corresponding physical coordinates, respectively.

[0049] In this embodiment, the panoramic overhead view is segmented by using a predetermined deep network model to segment the panoramic overhead view to obtain a pixel area of the parking space line, a pixel area of the lane line, and a pixel area of the column, and the semantic information is further processed by performing morphological operation on the pixel area to fill the pixel holes in the pixel area, thereby improving the robustness of subsequent processing and avoiding data errors or failure in subsequent image processing due to pixel holes. Further, the pixel coordinates of the parking space, the pixel coordinates of each lane line, and the pixel coordinates of each column are obtained by image processing on the pixel area of the parking space line, the pixel area of the lane line, and the pixel area of the column. Finally, the corresponding pixel coordinates are converted into corresponding physical coordinates in combination with the external parameters of the vehicle-mounted camera 1. The overall process is simple and efficient.

[0050] In specific implementation, the predetermined deep network model can use a U-net, DeepLab, or SegNe network model. In this embodiment, a U-net network model is used. When the U-net network model is pre-trained, the training pictures are first cropped to a uniform size, and the classes in the cropped images are manually labeled. The number of channels in the last layer of the U-net network model is adjusted to adjust the number of classes required for semantic segmentation. After obtaining the trained network model, the network model is used to perform semantic segmentation on the panoramic overhead view. In addition, it can be understood that when performing semantic segmentation, the lane ground markings and drivable free area markings in the panoramic overhead view are also extracted, and the above markings are also involved in constructing the final target global scene map. The lane ground markings and drivable free area markings in the constructed target global scene map can assist in path planning when the automatic parking system is applied.

[0051] In an optional embodiment of the present application, the pixel coordinates of the parking space are obtained by sequentially using a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the pixel area of the parking space line after morphological operation to obtain a plurality of parking space lines, and based on the relative positional relationship of each parking space line, the parking space and its pixel coordinates are obtained in the panoramic overhead view. In this embodiment, when the pixel area of the parking space line is processed, straight line detection and straight line clustering are sequentially performed, thereby greatly reducing the interference lines and reducing the data processing amount. After the parking space line is determined, the intersection of two parking space lines is a corner point of the parking space, and a plurality of parking spaces are quickly generated according to the relative positional relationship of each parking space line, and the pixel coordinates of the parking space in the panoramic overhead view are determined.

[0052] In an optional embodiment of the present application, the pixel coordinates of the lane lines are obtained by the following steps: sequentially adopting a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the pixel region of the lane lines after the morphological operation to obtain a plurality of suspected lines, and excluding the interfering lines with a length less than a preset length threshold and a horizontal distribution from each of the suspected lines to obtain each lane line in the panoramic view and the pixel coordinates thereof. In this embodiment, for the lane lines, which are also generally straight lines in the actual scene, the straight line detection and the straight line clustering are sequentially performed, so that the interfering lines can be greatly reduced, the data processing amount is reduced, and after a plurality of suspected lines are determined, the interfering lines with a certain length and not horizontally distributed are excluded, so that the remaining suspected lines are determined as the lane lines, and the pixel coordinates of the lane lines in the panoramic view are determined.

[0053] In an optional embodiment of the present application, the pixel coordinates of the column are obtained by the following steps: calculating the shortest pixel distance between each pixel point in the pixel region of the column and the parking space line and the lane line, removing the pixel points in the pixel region of the column with the shortest pixel distance greater than a preset pixel distance, sequentially adopting a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the remaining pixel points in the pixel region of the column to obtain the connecting line of the column and the parking ground, and constructing the column and the pixel coordinates thereof in the panoramic view based on the connecting line. In this embodiment, for the column, since the connecting line of the column and the parking ground is located on the ground together with the parking space line and the lane line, the shortest pixel distance between the pixel points on the connecting line and the parking space line and the lane line is generally not more than the preset pixel distance, based on which the pixel points on the connecting line are screened out, and then the straight line detection and the straight line clustering are adopted, so that the connecting line and the pixel coordinates thereof are obtained, and after the connecting line and the pixel coordinates thereof are determined, the column and the pixel coordinates thereof are determined according to a plurality of different connecting lines belonging to the column.

[0054] In the implementation, the predetermined straight line detection algorithm model mentioned in each of the foregoing embodiments can be a Hough straight line detection, an LSD straight line detection algorithm model, etc., and the Hough straight line detection algorithm model is selected in the embodiments of the present application, and the straight line clustering algorithm model can be a K-means algorithm, and the clustering method adopted in the embodiments of the present application is as follows:

[0055] First step: sorting the starting point pixel coordinates of the line segments from large to small according to the y value;

[0056] Second step: sequentially calculating the length of each line segment, and removing the line segment with a length less than a preset length;

[0057] Third step: calculate the slope of each remaining line segment in turn, compare two line segments with a slope difference less than a preset difference threshold, if the relative distance of the start point or end point of the two line segments is less than a preset distance, connect the start point of the first line segment and the end point of the second line segment to form a new line segment;

[0058] Fourth step: loop the third step to traverse all line segments, and finally obtain the line segment as the clustered line segment.

[0059] In an optional embodiment of the present application, the step S4 specifically comprises:

[0060] excluding the remaining physical coordinates in the global scene map belonging to the same parking space and except the latest physical coordinates;

[0061] excluding the remaining physical coordinates in the global scene map belonging to the same column and except the latest physical coordinates; and

[0062] merging the physical coordinates in the global scene map belonging to the same lane line by using a predetermined straight line clustering algorithm model.

[0063] In the embodiment, for the physical coordinates of the parking space and the column, only the latest physical coordinates (i.e. the physical coordinates obtained by the last time through image calculation) are retained, and for each line segment of the same lane line in the global scene map, the line segments are merged by using a predetermined straight line clustering algorithm model (which can be merged into one or a smaller number of line segments), thereby reducing the data processing amount.

[0064] In an optional embodiment of the present application, after excluding the repeated physical coordinates of the same predetermined target in the initial global scene map, the image constraint is further performed on each parking space in the initial global scene map to obtain the target global scene map, and the image constraint includes the same size constraint and lane line alignment constraint. Since the length and width of the parking space in the same parking lot should be consistent, and the side of the adjacent or side-by-side parking space should be aligned with the lane line, based on this, in the embodiment, the same size constraint and lane line alignment constraint are further performed on each parking space in the global scene map, thereby improving the position accuracy of the parking space.

[0065] In an optional embodiment of the present application, after constructing the local scene map of the motor vehicle at the current position coordinate based on the physical coordinates, the use state of the parking space in the local scene map is marked, and the use state of the parking space is mapped to the target global scene map, and each time the target global scene map is used for auxiliary driving, the use state of the parking space in the target global scene map is also updated, wherein the method for marking the use state of the parking space in the local scene map and updating the use state of the parking space in the target global scene map comprises:

[0066] identify the obstacle vehicles around the motor vehicle from each of the video images based on a predetermined vehicle recognition model (e.g., a yolov5 algorithm model);

[0067] calculate coordinate data of the obstacle vehicles in the global map coordinate system; and

[0068] mark or update the use state of the parking space according to the coordinate data of the obstacle vehicles.

[0069] In the embodiment, in the process of constructing the local scene map and the process of using the target global scene map subsequently, the use state of the parking space is also marked or updated, so that when the target global scene map is applied, the system can accurately determine whether the parking space is occupied in a short period of time, and thus flexibly select an idle parking space for parking.

[0070] In another aspect, an optional embodiment of the present application provides a motor vehicle parking lot scene map construction device 3 connected with a driving system 5 of a motor vehicle and each vehicle-mounted camera 1, the device 3 comprising a processor 30, a memory 32, and a computer program stored in the memory 32 and configured to be executed by the processor 30, the processor 30 implementing the motor vehicle parking lot scene map construction method according to any one of the above when executing the computer program.

[0071] For example, the computer program can be divided into one or more modules / units, which are stored in the memory 32 and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the motor vehicle parking lot scene map construction device 3. For example, the computer program can be divided into Figure 3 The motor vehicle parking lot scene map construction device 3 comprises a motor vehicle control module 41, an image processing module 42, a local map construction module 43, an initial global map construction module 44, and a target global map construction module 45, which correspond to steps S1-S5 above, respectively.

[0072] The motor vehicle parking lot scene map construction device 3 can be a desktop computer, a notebook, a palm computer, a cloud server, or the like. The motor vehicle parking lot scene map construction device 3 can include, but is not limited to, a processor 30 and a memory 32. Those skilled in the art can understand that the schematic diagram is only an example of the motor vehicle parking lot scene map construction device 3, and does not constitute a limitation on the motor vehicle parking lot scene map construction device 3, and can include more or fewer components than the diagram, or combine certain components, or different components, for example, the motor vehicle parking lot scene map construction device 3 can also include an input / output device, a network access device, a bus, and the like.

[0073] The processor 30 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic components, discrete hardware components, or the like. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, and the like. The processor 30 is the control center of the motor vehicle parking lot scene map construction device 3, and is connected to various parts of the motor vehicle parking lot scene map construction device 3 through various interfaces and lines.

[0074] The memory 32 can be used to store computer programs and / or modules, and the processor 30 realizes various functions of the motor vehicle parking lot scene map construction device 3 by running or executing computer programs and / or modules stored in the memory 32, and calling data stored in the memory 32. The memory 32 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function (such as a graphic recognition function, a graphic layering function, and the like), and the like; and the data storage area can store data (such as graphic data, and the like) created according to the use of the control device, and the like. In addition, the memory 32 can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0075] If the functions described in the embodiments of the present application are realized in the form of software function modules or units and sold or used as an independent product, they can be stored in a computer readable storage medium. Based on such an understanding, all or part of the flow of the method in the above-described embodiments can also be implemented by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program is executed by the processor 30, and the steps of each method embodiment described above can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files or some intermediate forms. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the contents of the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0076] In still another aspect, an optional embodiment of the present application provides a computer readable storage medium, including a stored computer program, wherein the computer program controls the device where the computer readable storage medium is located to perform the motor vehicle parking lot scene map construction method according to any one of the above-described embodiments when the computer program is running.

[0077] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other.

[0078] The embodiments of the present application are described above in combination with the drawings, but the present application is not limited to the above-described specific embodiments. The above-described specific embodiments are only illustrative, but not restrictive. Those skilled in the art can make many forms under the inspiration of the present application without departing from the purpose of the present application and the scope protected by the claims, which are all within the protection scope of the present application.

Claims

1. A method of constructing a map of a parking lot scene for motor vehicles, characterized in that The method comprises the following steps: constructing a global map coordinate system with a preset starting point of the parking lot as the coordinate origin, and controlling the motor vehicle to travel from the preset starting point to a preset terminal point according to a preset travel path; in the process of motor vehicle travel, controlling each vehicle-mounted camera of the motor vehicle to collect video images of the surrounding environment of the motor vehicle in real time, and generating a panoramic overhead view based on the video images collected by each vehicle-mounted camera; acquiring the current position coordinates of the motor vehicle in the global map coordinate system at a predetermined period and constructing a local scene map based on the current position coordinates, the construction process of the local scene map comprising: extracting semantic information of predetermined targets in the parking lot from the panoramic overhead view, processing the semantic information to obtain physical coordinates of the predetermined targets having actual physical lengths, and constructing a local scene map of the motor vehicle at the current position coordinates based on the physical coordinates, the predetermined targets at least including parking spaces, lane lines, and pillars; mapping the local scene maps constructed in each period in the global map coordinate system according to the current position coordinates obtained in the corresponding period to obtain an initial global scene map; and excluding the physical coordinates of the same predetermined targets that are repeated in the initial global scene map to obtain a target global scene map of the parking lot; wherein the semantic information of the predetermined targets refers to pixel regions of parking space lines, pixel regions of lane lines, and pixel regions of pillars obtained by performing semantic segmentation on the panoramic overhead view using a predetermined deep network model, and the processing of the semantic information to obtain the physical coordinates of the predetermined targets having actual physical lengths specifically comprises: performing morphological operations on the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars respectively to fill pixel holes in the pixel regions; performing image processing on the pixel regions of the parking space lines, the pixel regions of the lane lines, and the pixel regions of the pillars after morphological operations to obtain pixel coordinates of each parking space, pixel coordinates of each lane line, and pixel coordinates of each pillar in the panoramic overhead view, respectively; and converting the pixel coordinates of each parking space, the pixel coordinates of each lane line, and the pixel coordinates of each pillar in the panoramic overhead view into corresponding physical coordinates, respectively.

2. The motor-park scenario map construction method according to claim 1, wherein The pixel coordinates of the parking spaces are obtained by: sequentially using a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the pixel regions of the parking space lines after morphological operations to obtain a plurality of parking space lines, and constructing a parking space and its pixel coordinates in the panoramic overhead view based on the relative positional relationship of each parking space line.

3. The motor-park scenario map construction method according to claim 1, wherein The pixel coordinates of the lane lines are obtained by: sequentially using a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the pixel regions of the lane lines after morphological operations to obtain a plurality of suspected lines, and excluding interference lines having a length less than a preset length threshold and a horizontal distribution from each suspected line to obtain each lane line and its pixel coordinates in the panoramic overhead view.

4. The motor vehicle parking lot map construction method according to claim 1 or 2 or 3, characterized by, The pixel coordinates of the column are obtained by the following steps: calculating the shortest pixel distance between each pixel point in the pixel area of the column and the parking space line and the lane line, removing the pixel points in the pixel area of the column whose shortest pixel distance is greater than a preset pixel distance, sequentially using a predetermined straight line detection algorithm model and a predetermined straight line clustering algorithm model to process the remaining pixel points in the pixel area of the column to obtain a connecting line of the column and the parking lot ground, and constructing the column and its pixel coordinates in the panoramic overhead view based on the connecting line.

5. The motor-park scenario map construction method according to claim 1, wherein The excluding the same predetermined target from the initial global scene map includes: excluding the remaining physical coordinates in the global scene map that belong to the same parking space and are except the latest physical coordinates; excluding the remaining physical coordinates in the global scene map that belong to the same column and are except the latest physical coordinates; and merging the physical coordinates in the global scene map that belong to the same lane line using a predetermined straight line clustering algorithm model.

6. The motor vehicle parking lot map construction method according to claim 1 or 5, characterized by, After excluding the same predetermined target from the initial global scene map, the image constraint of each parking space in the initial global scene map is performed to obtain the target global scene map, and the image constraint includes the same size constraint and lane line alignment constraint.

7. The motor-park scenario map construction method according to claim 1, wherein After constructing the local scene map of the motor vehicle at the current position coordinates based on the physical coordinates, the use state of the parking space in the local scene map is marked, and the use state of the parking space is mapped to the target global scene map. Each time the target global scene map is used for auxiliary driving, the use state of the parking space in the target global scene map is also updated, wherein the method for marking the use state of the parking space in the local scene map and updating the use state of the parking space in the target global scene map includes: identifying the obstacle vehicles around the motor vehicle from each video image based on a predetermined vehicle recognition model; calculating the coordinate data of the obstacle vehicles in the global map coordinate system; and marking or updating the use state of the parking space according to the coordinate data of the obstacle vehicles.

8. A motor vehicle parking lot scene map construction device connected to a driving system of a motor vehicle and each vehicle-mounted camera, characterized by, The device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and the processor implements the motor vehicle parking lot scene map construction method according to any one of claims 1 to 7 when executing the computer program.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium includes a stored computer program, wherein the computer readable storage medium controls the device where the computer readable storage medium is located to execute the motor vehicle parking lot scene map construction method according to any one of claims 1 to 7 when the computer program runs.

Images