A positioning method, device, equipment and medium for an automobile

Abstract

The present invention provides a method, apparatus, device, and medium for locating a vehicle. The method comprises acquiring multiple raw image data collected from a target area; statistically processing the multiple raw image data to generate positioning map data for the target area; acquiring road image data of the current vehicle location, wherein the current vehicle is located within the target area; generating semantic feature data based on the road image data, wherein the semantic feature data represents ground elements at the current vehicle location and the number of such ground elements; and matching the semantic feature data with the positioning map data to generate the current vehicle location data. The present invention can achieve low-cost positioning of vehicles in urban areas, and the positioning method has high execution efficiency and high stability.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous vehicle positioning, and in particular to a method, device, equipment and medium for positioning a vehicle. Background Art

[0002] Vehicle positioning is essential for autonomous driving. However, in urban environments, GNSS (Global Navigation Satellite System) signals often fail due to obstructions such as tall buildings, trees, and tunnels. Furthermore, traditional visual feature positioning methods are affected by season, weather, lighting, viewing angle, and occlusion. Accurate and robust positioning based on visual features over long distances is difficult to achieve, necessitating the use of other complementary positioning methods. Summary of the Invention

[0003] In view of the above shortcomings of the prior art, the present invention provides a method, device, equipment and medium for positioning an automobile to solve the above technical problems.

[0004] The present invention provides a method for positioning a car, comprising:

[0005] Acquire multiple original image data collected from the target area;

[0006] Performing statistical processing on the plurality of original image data to generate positioning map data of the target area;

[0007] Obtaining road image data of a current vehicle location, wherein the current vehicle is located within the target area;

[0008] generating semantic feature data according to the road image data, wherein the semantic feature data represents the ground elements at the current position of the vehicle and the number of the ground elements; and

[0009] The semantic feature data and the positioning map data are matched to generate the current vehicle position data.

[0010] In one embodiment of the present invention, the step of matching the semantic feature data with the positioning map data to generate the current vehicle location data includes:

[0011] performing similarity calculation processing on the positioning map data and the semantic feature data to generate edit distance data; and

[0012] The current vehicle position data is obtained according to the edit distance data.

[0013] In one embodiment of the present invention, the step of performing similarity calculation on the positioning map data and the semantic feature data to generate edit distance data includes:

[0014] Converting the positioning map data into first character string data and converting the semantic feature data into second character string data; and

[0015] A distance calculation process is performed on the first character string data and the second character string data to generate the edit distance data.

[0016] In one embodiment of the present invention, the distance calculation process satisfies the following formula:

[0017]

[0018] Wherein: edit[i][j] represents the edit distance data, i represents the i-th position of the first string, j represents the j-th position of the second string, min function represents the shortest search path from the first string to the second string, and flag represents the number of valid replacements of the mark.

[0019] In one embodiment of the present invention, the step of performing statistical processing on the plurality of original image data to generate positioning map data of the target area includes:

[0020] Segmenting the plurality of original image data to generate semantic label information, wherein the semantic label information represents a plurality of static elements and a plurality of dynamic elements in the original image data; and

[0021] The positioning map data is generated according to the semantic tag information.

[0022] In one embodiment of the present invention, the step of generating the positioning map data according to the semantic tag information includes:

[0023] Extracting semantic element information from the original image data based on the semantic tag information, where the semantic element information represents lane lines, arrows, stop lines, and zebra crossings in the original image data; and

[0024] Aggregation processing is performed on the semantic element information to generate the positioning map data.

[0025] In one embodiment of the present invention, the step of generating semantic feature data based on the road image data includes:

[0026] performing data compression processing on the road image data to generate road information; and

[0027] Clustering is performed on the road information to generate the semantic feature data.

[0028] The present invention also provides a vehicle positioning device, comprising:

[0029] A map acquisition module is used to obtain multiple original image data collected from the target area;

[0030] A map generation module, configured to perform statistical processing on the plurality of original image data to generate positioning map data of the target area;

[0031] A road perception module is used to obtain road image data of a current vehicle location, where the current vehicle is located in the target area;

[0032] a feature extraction module, configured to generate semantic feature data based on the road image data, wherein the semantic feature data represents the ground elements at the current location of the vehicle and the number of the ground elements; and

[0033] The matching and positioning module is used to match the semantic feature data with the positioning map data to generate the current vehicle position data.

[0034] The present invention further provides an electronic device, comprising:

[0035] one or more processors;

[0036] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the electronic device implements the automobile positioning method as described in any one of the above.

[0037] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor of a computer, the computer is caused to execute any one of the above-mentioned methods for positioning a vehicle.

[0038] Beneficial effects of the present invention: The present invention provides a positioning method based on positioning map data and semantic feature data, which can realize the positioning of a car in an urban area at a low cost, and the positioning method has high execution efficiency, high result stability, and high use value.

[0039] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present invention, and together with the specification, are used to explain the principles of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and it is clear that a person skilled in the art can derive other drawings based on these drawings without inventive effort. In the drawings:

[0041] Figure 1 This is an application environment of a vehicle positioning method provided by the present invention;

[0042] Figure 2 This is a flow chart of a method for positioning a vehicle provided by the present invention;

[0043] Figure 3 yes Figure 2 A schematic flow chart of a specific implementation of step S202;

[0044] Figure 4 yes Figure 3 A schematic flow chart of a specific implementation of step S302;

[0045] Figure 5 yes Figure 2 A schematic flow chart of a specific implementation of step S204;

[0046] Figure 6 yes Figure 5 Schematic diagram of the process of step S501;

[0047] Figure 7 yes Figure 2 A schematic flow chart of a specific implementation of step S205;

[0048] Figure 8 yes Figure 6 A schematic flow chart of a specific implementation of step S502;

[0049] Figure 9 This is a structural block diagram of a vehicle positioning device provided by the present invention;

[0050] Figure 10 It is a schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following describes the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art will readily appreciate the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the various details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are intended only to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

[0052] The figures provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Thus, the figures only show components relevant to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be arbitrarily changed, and the component layout may also be more complex.

[0053] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the embodiments of the present invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring the embodiments of the present invention.

[0054] The present invention proposes a method for positioning a car, which can be applied in the following situations: Figure 1 In an application environment, client 101 can communicate with server 102 via a network. Server 102 collects data about a target area using multiple vehicles to construct map data for the target area. Client 101 uses the vehicle's front-facing camera to obtain road image data at the vehicle's current location and uses this road image data to match the map data to complete vehicle positioning. The present invention is described in detail below using specific embodiments.

[0055] See also Figure 2 As shown, Figure 2 The figure is a flow chart of a method for positioning a vehicle provided by the present invention, which may include the following steps:

[0056] Step S201: Acquire a plurality of original image data collected from a target area.

[0057] In one embodiment of the present invention, when step S201 is executed, multiple raw image data of the target area are acquired. Specifically, step S201 acquires multiple raw image data of the target area, where the raw image data can be acquired using multiple vehicles with environmental perception capabilities. In this embodiment, the environmental perception devices of the vehicles can be onboard forward-looking cameras.

[0058] It should be noted that the high-precision map data used for car positioning includes lane markings, stop lines, crosswalks, traffic signs, traffic lights and other existing traffic facilities on the road surface and roadside, as well as the connectivity between roads, the horizontal and vertical correlation between lanes, and the correlation between traffic facilities and roads / lanes, etc., which are used for path planning and driving decision-making. In order to generate the above information, the existing technology usually adopts centralized surveying and mapping means, and performs map surveying and mapping by centrally deploying special map collection vehicles. The collection cost of centralized surveying and mapping means is high, and the production cycle is long. At the same time, the topological connectivity relationship between roads at intersections is often inconsistent with the actual situation. The present invention adopts a crowdsourcing method for map surveying and mapping. By using multiple cars with environmental perception capabilities, the cars collect road image data while driving and upload them to the cloud. The cloud constructs a positioning map data map with high restoration and real-time update based on the feedback data.

[0059] Step S202: Statistically process the plurality of original image data to generate positioning map data of the target area.

[0060] In one embodiment of the present invention, when step S202 is executed, statistical processing is performed on multiple raw image data to generate positioning map data for the target area. Specifically, step S202 may first perform semantic segmentation processing on each frame of the raw image data based on a deep learning algorithm to generate semantic label information for each pixel in the target area. The semantic label information represents multiple static and dynamic elements in the raw image data, and may include pedestrians, cars, sidewalks, traffic lights, etc. The raw image data is then processed based on the semantic label information to generate positioning map data for the target area.

[0061] It should be noted that the positioning map data contains multiple module locations, each of which has its corresponding semantic cluster feature descriptor. The semantic feature cluster descriptor represents a form of representation of the surrounding point cloud categories and distribution centered on the semantic object, referring to the orientation of the semantic object and the surrounding pixel information. In this embodiment, the positioning map data does not require data processing of spatial elements such as signboards, only road surface information such as lane lines, arrows, stop lines, zebra crossings, etc.

[0062] Step S203: Obtain road image data of the current vehicle location, and the current vehicle is located in the target area.

[0063] In one embodiment of the present invention, when step S203 is executed, road image data of the current vehicle's location is obtained, indicating that the current vehicle is within the target area. Specifically, step S203 may obtain road image data of the current vehicle's location using a sensor onboard the current vehicle, indicating that the current vehicle is within the target area. In this embodiment, the sensor onboard the current vehicle may be a structured light camera, a binocular vision camera, or a monocular vision camera.

[0064] Step S204: Generate semantic feature data based on the road image data, where the semantic feature data represents the ground elements and the number of ground elements at the current location of the car.

[0065] In one embodiment of the present invention, when step S204 is executed, semantic feature data is generated based on the road image data, and the semantic feature data represents the ground elements and the number of ground elements at the current location of the car. Specifically, step S204 may first generate the road information of the current location based on the road image data obtained by the current car, and then generate the semantic feature information of its location. The semantic feature data represents the ground elements and the number of ground elements at the current location of the car. Semantic feature information is also called a semantic feature cluster descriptor, and its specific form is the semantic type of the ground element and the corresponding number of objects. For example, the current road information contains X lane lines and Y arrows.

[0066] Step S205: Match the semantic feature data with the positioning map data to generate the current vehicle location data.

[0067] In one embodiment of the present invention, when step S205 is executed, the semantic feature data and the positioning map data are matched to generate the current vehicle location data. Specifically, step S205 generates the current vehicle location data by matching the semantic feature data of the current vehicle location with multiple module locations in the positioning map data. The matching method can be to calculate the edit distance between the semantic feature data and the semantic clustering feature descriptors corresponding to the multiple module locations. The module location with the smallest calculated edit distance value is the current vehicle location.

[0068] See also Figure 3 As shown, Figure 3 yes Figure 2 A flow chart of a specific implementation of step S202 is shown in FIG. Step S202 may include the following steps:

[0069] Step S301 : Segment a plurality of original image data to generate semantic label information, where the semantic label information represents a plurality of static elements and a plurality of dynamic elements in the original image data.

[0070] In one embodiment of the present invention, when step S301 is executed, multiple original image data are segmented to generate semantic label information, and the semantic label information represents multiple static elements and multiple dynamic elements in the original image data. Specifically, based on the deep learning algorithm, semantic segmentation is performed on each frame of the original image data to generate corresponding semantic image data. The pixel points of the above-mentioned semantic image data contain semantic label information. The semantic label information represents multiple static elements and multiple dynamic elements in the original image data, which may include pedestrians, cars, sidewalks, traffic lights, etc. In this embodiment, the deep learning algorithm may be a convolutional neural network algorithm. The segmentation algorithm based on the convolutional neural network algorithm is an algorithm that performs binary classification on each pixel, which is mainly divided into semantic segmentation and instance segmentation. Among them, semantic segmentation is at the pixel level, and each pixel in the image is divided into a corresponding category, that is, pixel-level classification is achieved.

[0071] Step S302: Generate positioning map data based on the semantic tag information.

[0072] In one embodiment of the present invention, when step S302 is executed, positioning map data is generated based on the semantic tag information. Specifically, semantic element information is first extracted from the image data based on the semantic tag information. The semantic element information is then aggregated to generate the positioning map data.

[0073] See also Figure 4 As shown, Figure 4 yes Figure 3 A flow chart of a specific implementation of step S302 is shown in FIG. Step S302 may include the following steps:

[0074] Step S401: extracting semantic element information from the original image data based on the semantic tag information, where the semantic element information represents lane lines, arrows, stop lines, and zebra crossings in the original image data.

[0075] In one embodiment of the present invention, when step S401 is executed, semantic element information is extracted from the original image data based on the semantic label information. The semantic element information represents lane lines, arrows, stop lines, and zebra crossings in the original image data. Specifically, based on the semantic label information contained in the pixels in the semantic image data, the semantic element information required for map construction can be extracted from the semantic image data. In this embodiment, the semantic element information may include stable ground semantic elements such as lane lines, arrows, stop lines, and zebra crossings.

[0076] Step S402: Aggregate the semantic element information to generate positioning map data.

[0077] In one embodiment of the present invention, when step S402 is executed, the semantic element information is aggregated to generate positioning map data. Specifically, point cloud data for each frame is first generated based on the semantic element information extracted in step S401. The point cloud data corresponding to all frames are then spliced together to generate complete positioning map data for the target area. The positioning map data is configured with multiple module locations, each of which has a corresponding semantic cluster feature descriptor.

[0078] See also Figure 5 As shown, Figure 5 yes Figure 2 A flow chart of a specific implementation of step S204 is shown in FIG. Step S204 may include the following steps:

[0079] Step S501: compress the road image data to generate road information.

[0080] In one embodiment of the present invention, when step S501 is executed, data compression processing is performed on the road image data to generate road information. Specifically, the perception data obtained by the current vehicle through the front-view camera can be the original road image. In the data compression processing, semantic segmentation processing is first performed on the original road image to obtain the corresponding road semantic image. The semantic image is then processed to generate point cloud data of the current vehicle location. The main purpose of generating point cloud data is to remove non-ground elements such as pedestrians and vehicles. Finally, the point cloud data is converted into an image to generate road information. Due to the large amount of data in each frame of point cloud data, direct clustering calculation takes a long time, and for the descriptors used in the calculation, a large amount of point cloud data has information redundancy. Therefore, in order to improve computing efficiency, the point cloud data of the current vehicle location is first converted into an image. In this embodiment, the road information and point cloud data are stored using the data types provided by OpenCV and PCL, respectively. In addition, to further improve computing efficiency, the converted road information can also be processed with reduced resolution to reduce the amount of data.

[0081] See also Figure 6 As shown, Figure 6 yes Figure 5 Figure 5 shows the process of step S501 in Figure 5. In one embodiment of the present invention, first, step S601 processes the original road image to generate point cloud data for the current vehicle location. The point cloud data contains 2113 points. Then, step S602 converts the point cloud data to generate road information. The road information image has a pixel size of 198*272. Finally, step S603 performs multiple downsampling operations on the road information. After multiple downsampling, the resolution of the road information is reduced from 198*272 to 24*34.

[0082] Step S502: clustering the road information to generate semantic feature data.

[0083] In one embodiment of the present invention, when step S502 is executed, the road information is clustered to generate semantic feature data. Specifically, step S502 uses a density-based clustering method to cluster the road information to generate semantic feature data of the current frame, i.e., a semantic feature cluster descriptor.

[0084] Furthermore, different semantic features are optimized based on their contribution to the effectiveness of positioning and their different proportions in the semantic feature data can improve the accuracy of positioning.

[0085] See also Figure 7 As shown, Figure 7 yes Figure 2 A flow chart of a specific implementation of step S205 is shown in FIG. Step S205 may include the following steps:

[0086] Step S701: perform similarity calculation on the positioning map data and the semantic feature data to generate edit distance data.

[0087] In one embodiment of the present invention, when step S701 is executed, similarity calculation processing is performed on the positioning map data and the semantic feature data to generate edit distance data. Specifically, the semantic feature data generated in step S204 and the positioning map data generated in step S202 are subjected to similarity calculation processing to realize the positioning of the current car. The above-mentioned similarity calculation processing needs to be performed based on the edit distance between the descriptor of the current car location and the module location set in the positioning map data. Step S701 first converts the semantic feature data of the current car location and the semantic clustering feature descriptor of the module location in the positioning map data into a first string and a second string respectively. Then the edit distance between the first string and the second string is calculated.

[0088] Edit distance, also known as Levenshtein distance, is a quantitative measure of the difference between two strings. It's measured by calculating the minimum number of operations required to transform one string into the other. Specifically, it measures the minimum number of character operations required to transform string A into string B. These operations include deleting a character, inserting a character, and modifying a character. For example, for the strings 'if' and 'iff', inserting an 'f' or deleting an 'f' can achieve this. Generally speaking, the smaller the edit distance between two strings, the more similar they are. If two strings are equal, their edit distance is 0, meaning no operations are required.

[0089] Step S702: Obtain the current vehicle location data based on the edit distance data.

[0090] In one embodiment of the present invention, when step S702 is executed, the current vehicle location data is obtained based on the edit distance data. Specifically, among multiple module locations, the module location with the minimum edit distance to the current vehicle location is found, and the location coordinates of this location are considered to be the current vehicle location data.

[0091] See also Figure 8 As shown, Figure 8 yes Figure 7 A flow chart of a specific implementation of step S702 is shown in FIG. Step S702 may include the following steps:

[0092] Step S801: Convert positioning map data into first character string data, and convert semantic feature data into second character string data.

[0093] In one embodiment of the present invention, when step S801 is executed, the positioning map data is converted into a first string of data, and the semantic feature data is converted into a second string of data. Specifically, based on the number of semantic types, the semantic feature data of the current vehicle location and the semantic clustering feature descriptors of the module locations in the positioning map data are converted into the first string of data and the second string of data, respectively.

[0094] Step S802 : performing distance calculation processing on the first character string data and the second character string data to generate edit distance data.

[0095] In one embodiment of the present invention, when step S802 is executed, a distance calculation process is performed on the first string data and the second string data to generate edit distance data. Specifically, the edit distance between the first string data and the second string data is calculated to satisfy the following dynamic programming equation:

[0096]

[0097] Among them: edit[i][j] represents the edit distance data, that is, the shortest search path between the first string starting from the 0th character to the i-th string and the second string starting from the 0th character to the j-th string, i represents the i-th position of the first string, j represents the j-th position of the second string, min function represents the shortest search path from the first string to the second string, and flag represents the number of valid replacements of the mark.

[0098] In this embodiment, the value of Flag satisfies the following formula:

[0099]

[0100] Wherein, A represents the first character string, and B represents the second character string.

[0101] See also Figure 9 As shown, Figure 9 1 is a structural block diagram of an exemplary embodiment of the present invention showing a positioning device for an automobile. In some embodiments, the positioning device for an automobile includes a map acquisition module 901, a map generation module 902, a road perception module 903, a feature extraction module 904 and a matching positioning module 905. Each functional module is described in detail as follows.

[0102] The map acquisition module 901 is used to acquire a plurality of original image data collected from a target area.

[0103] In one embodiment of the present invention, the map acquisition module 901 may be used to acquire multiple raw image data of a target area. Specifically, the map acquisition module 901 acquires multiple raw image data of the target area. The raw image data may be acquired using multiple vehicles with environmental perception capabilities. In this embodiment, the environmental perception devices in the vehicles may be onboard forward-looking cameras.

[0104] The map generation module 902 is used to perform statistical processing on a plurality of original image data to generate positioning map data of a target area.

[0105] In one embodiment of the present invention, the map generation module 902 may be configured to statistically process multiple raw image data to generate positioning map data for the target area. Specifically, the map generation module 902 may first perform semantic segmentation on each frame of the raw image data based on a deep learning algorithm to generate semantic label information for each pixel in the target area. The semantic label information represents multiple static and dynamic elements in the raw image data, and may include pedestrians, cars, sidewalks, traffic lights, etc. The raw image data is then processed based on the semantic label information to generate positioning map data for the target area.

[0106] In a specific embodiment, the map generation module 902 may also be specifically configured to:

[0107] Segmenting the plurality of original image data to generate semantic label information, where the semantic label information represents a plurality of static elements and a plurality of dynamic elements in the original image data; and

[0108] Generate positioning map data based on semantic label information.

[0109] The road perception module 903 is used to obtain road image data of the current position of the car, and the current car is located in the target area.

[0110] In one embodiment of the present invention, the road perception module 903 may be configured to obtain road image data at the current vehicle's location, indicating that the vehicle is within a target area. Specifically, the road perception module 903 may obtain road image data at the vehicle's location using a sensor installed on the vehicle, indicating that the vehicle is within the target area. In this embodiment, the sensor installed on the vehicle may be a structured light camera, a binocular camera, or a monocular camera.

[0111] The feature extraction module 904 is used to generate semantic feature data based on the road image data. The semantic feature data represents the ground elements and the number of ground elements at the current position of the car.

[0112] In one embodiment of the present invention, the feature extraction module 904 may be used to generate semantic feature data based on road image data, where the semantic feature data represents the ground elements and the number of ground elements at the current vehicle location. Specifically, the feature extraction module 904 may first generate road information at the current vehicle location based on the road image data acquired by the current vehicle, and then generate semantic feature information at its location. The semantic feature data represents the ground elements and the number of ground elements at the current vehicle location. Semantic feature information is also called a semantic feature cluster descriptor, and its specific form is the semantic type of the ground element and the corresponding number of objects, for example, the current road information contains X lane lines and Y arrows.

[0113] In a specific embodiment, the feature extraction module 904 may also be specifically used to:

[0114] performing data compression processing on the road image data to generate road information; and

[0115] The road information is clustered to generate semantic feature data.

[0116] The matching and positioning module 905 is used to match the semantic feature data with the positioning map data to generate the current position data of the vehicle.

[0117] In one embodiment of the present invention, the matching and positioning module 905 can be used to match semantic feature data with positioning map data to generate the current vehicle location data. Specifically, the matching and positioning module 905 generates the current vehicle location data by matching the semantic feature data of the current vehicle location with multiple module locations in the positioning map data. The matching method can be to calculate the edit distance between the semantic feature data and the semantic clustering feature descriptors corresponding to the multiple module locations. The module location with the smallest value in the calculated edit distance data is the current vehicle location.

[0118] In a specific embodiment, the matching and positioning module 905 may also be specifically configured to:

[0119] Performing similarity calculation on the positioning map data and the semantic feature data to generate edit distance data; and

[0120] Based on the edit distance data, obtain the current car position data.

[0121] It should be noted that the vehicle positioning device provided in the above-described embodiment and the vehicle positioning method provided in the above-described embodiment are based on the same concept. The specific manner in which the various modules and units perform their operations has been described in detail in the method embodiments and will not be repeated here. In actual applications, the vehicle positioning device provided in the above-described embodiment can, as needed, distribute the aforementioned functions among different functional modules, i.e., divide the internal structure of the system into different functional modules to perform all or part of the functions described above. This is not intended to be limiting herein.

[0122] An embodiment of the present invention further provides an electronic device comprising: one or more processors; and a storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the electronic device implements the vehicle positioning method provided in each of the above embodiments.

[0123] Please refer to Figure 10 As shown, Figure 10 This is a schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present invention. Figure 10 The computer system 1000 of the electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0124] like Figure 10 As shown, the computer system 1000 includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 1002 or the program loaded from the storage part 1008 into the random access memory (RAM) 1003, such as executing the method in the above embodiment. Various programs and data required for system operation are also stored in the RAM 1003. The CPU 1001, ROM 1002 and RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0125] The following components are connected to the I / O interface 1005: an input section 1006 including a keyboard, a mouse, and the like; an output section 1007 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 1008 including a hard disk and the like; and a communication section 1009 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as needed. Removable media 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 1010 as needed, so that computer programs read therefrom can be installed into the storage section 1008 as needed.

[0126] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product that includes a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 1009 and / or installed from the removable medium 1011. When the computer program is executed by the central processing unit (CPU) 1001, the various functions defined in the system of the present invention are performed.

[0127] It should be noted that the computer-readable medium shown in the embodiments of the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. The computer-readable storage medium may, for example, be an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable computer program. Such a propagated data signal may take a variety of forms, including, but not limited to, an electromagnetic signal, an optical signal, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0128] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the boxes can also occur in an order different from that marked in the accompanying drawings. For example, two boxes shown in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0129] The units involved in the embodiments of the present invention may be implemented in software or hardware, and the units described may also be provided in a processor. In some cases, the names of these units do not limit the units themselves.

[0130] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon. When executed by a computer processor, the computer program causes the computer to perform the aforementioned vehicle positioning method. The computer-readable storage medium may be included in the electronic device described in the above embodiments, or may exist independently and not be incorporated into the electronic device.

[0131] Another aspect of the present invention provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the vehicle positioning method provided in each of the above embodiments.

[0132] In summary, the present invention provides a method, system, device and medium for positioning a car. The present invention first generates crowdsourced positioning map data of a target area based on original image data. The map is mainly composed of point cloud data containing valid semantic element information of road elements; then, semantic feature data is generated based on the road image data acquired by the current vehicle, and the semantic feature data is used to match the crowdsourced positioning map data of the urban area of the target area to complete the positioning. The positioning map data is generated in a crowdsourcing manner, and its data collection is convenient and low-cost. The semantic feature data adopts a method of converting point cloud data into images and then compressing and clustering them, which improves the computing efficiency. In addition, the edit distance is used to represent the distance between the module position of the positioning map data and the semantic feature data, which facilitates position matching. The positioning method provided by the present invention has high execution efficiency, high result stability, and high use value.

[0133] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, any equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical principles disclosed herein are intended to be covered by the claims of the present invention.

Claims

1. A method for positioning a car, characterized in that: include: Acquire multiple original image data collected from the target area; Performing statistical processing on the plurality of original image data to generate positioning map data of the target area; Obtaining road image data of a current vehicle location, wherein the current vehicle is located within the target area; Generating semantic feature data according to the road image data, wherein the semantic feature data represents the ground elements at the current position of the vehicle and the number of the ground elements; as well as Performing similarity calculation on the positioning map data and the semantic feature data to generate edit distance data; as well as The current vehicle position data is obtained according to the edit distance data.

2. The method for positioning a vehicle according to claim 1, wherein: The step of performing similarity calculation on the positioning map data and the semantic feature data to generate edit distance data includes: Converting the positioning map data into first character string data and converting the semantic feature data into second character string data; and A distance calculation process is performed on the first character string data and the second character string data to generate the edit distance data.

3. The method for positioning a vehicle according to claim 2, wherein: The distance calculation process satisfies the following formula: Wherein: edit[i][j] represents the edit distance data, i represents the i-th position of the first string, j represents the j-th position of the second string, min function represents the shortest search path from the first string to the second string, and flag represents the number of valid replacements of the mark.

4. The method for positioning a vehicle according to claim 1, wherein: The step of performing statistical processing on the plurality of original image data to generate positioning map data of the target area includes: Segmenting the plurality of original image data to generate semantic label information, wherein the semantic label information represents a plurality of static elements and a plurality of dynamic elements in the original image data; and The positioning map data is generated according to the semantic tag information.

5. The method for positioning a vehicle according to claim 4, wherein: The step of generating the positioning map data according to the semantic tag information includes: Extracting semantic element information from the original image data based on the semantic tag information, where the semantic element information represents lane lines, arrows, stop lines, and zebra crossings in the original image data; and Aggregation processing is performed on the semantic element information to generate the positioning map data.

6. The method for positioning a vehicle according to claim 1, wherein: The step of generating semantic feature data based on the road image data includes: performing data compression processing on the road image data to generate road information; and Clustering is performed on the road information to generate the semantic feature data.

7. A positioning device for a car, characterized in that: include: A map acquisition module is used to obtain multiple original image data collected from the target area; A map generation module, configured to perform statistical processing on the plurality of original image data to generate positioning map data of the target area; A road perception module is used to obtain road image data of a current vehicle location, where the current vehicle is located in the target area; a feature extraction module, configured to generate semantic feature data based on the road image data, wherein the semantic feature data represents the ground elements at the current position of the vehicle and the number of the ground elements; as well as The matching positioning module is used to perform similarity calculation processing on the positioning map data and the semantic feature data to generate edit distance data; and obtain the current vehicle position data based on the edit distance data.

8. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the automobile positioning method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by a processor of a computer, the computer is caused to execute the vehicle positioning method according to any one of claims 1 to 6.

Images