Grid map compression method and system, vehicle, electronic device and storage medium

Abstract

The invention discloses a grid map compression method and system, a vehicle, an electronic device and a storage medium. The grid map compression method comprises the steps of obtaining an initial grid map; performing down-sampling on the initial grid map to obtain a compressed map, and encoding features of the initial grid map to obtain a first feature code; and updating the initial grid map by using the compressed map and the first feature code. According to the technical scheme provided by the invention, the technical problem that the raster map cannot be used on some map systems with limited hardware in the prior art can be solved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a raster map compression method, system, vehicle, electronic device and storage medium. Background Art

[0002] Occupancy grid maps can provide an effective method to express and process environmental information by converting spatial point coordinates into discrete grid forms, thereby expressing information such as the boundaries and locations of obstacles.

[0003] However, the use of raster maps has high hardware requirements, especially high memory requirements. Therefore, raster maps cannot be used on some map systems with limited hardware. Summary of the Invention

[0004] The present invention provides a raster map compression method, system, vehicle, electronic device and storage medium, aiming to effectively solve the technical problem in the prior art that raster maps cannot be used on some hardware-limited map systems.

[0005] According to a first aspect of the present invention, the present invention provides a raster map compression method, comprising: obtaining an initial raster map; downsampling the initial raster map to obtain a compressed map, and encoding features of the initial raster map to obtain a first feature code; and updating the initial raster map using the compressed map and the first feature code.

[0006] Furthermore, the step of downsampling the initial grid map includes: sampling the initial grid map using a first preset sampling window, wherein the first preset sampling window has a first predetermined number of grids; and projecting the map sampled each time by the first preset sampling window into an area in the order of sampling the initial grid map, wherein the area corresponding to the projection of the map sampled by each first preset sampling window includes a second predetermined number of grids, and the first predetermined number is greater than the second predetermined number.

[0007] Furthermore, the step of encoding the features of the initial grid map includes: encoding the position of each grid occupied feature in the initial grid map, and encoding the probability value of each grid occupied feature.

[0008] Furthermore, the step of position encoding the occupied features of each grid in the initial grid map includes: obtaining blank grids and non-blank grids in the grid map; performing a first mark on the blank grids and a second mark on the non-blank grids, wherein the first mark and the second mark are different.

[0009] Furthermore, the step of encoding the probability value of each grid being occupied by a feature includes: calculating the probability value of the map occupying each grid; and encoding the numerical value of the probability value as the probability value of the occupied feature.

[0010] Furthermore, after encoding the features of the initial grid map to obtain the first feature codes, the method further includes: storing the first feature codes in a predetermined number and in a predetermined format.

[0011] Furthermore, the method also includes: acquiring the position state of the target vehicle and the environmental point cloud data obtained by radar detection of the target vehicle in real time; mapping the environmental point cloud data into the grid map; re-encoding the grid mapped with the environmental point cloud data to obtain a second feature code; and using the second feature code to update the first feature code of the grid mapped with the environmental point cloud data.

[0012] Furthermore, the step of re-encoding the grid mapped by the environmental point cloud data includes: calculating the occupancy probability of the grid occupied by the point cloud data using the range and azimuth data of the environmental point cloud data.

[0013] Furthermore, in the calculation process of the occupancy probability, the occupancy probability is calculated using the normal distribution of the range and azimuth data of the environmental point cloud data, or the occupancy probability is calculated using the Gaussian distribution of the range and azimuth data of the environmental point cloud data.

[0014] According to a second aspect of the present invention, the present invention further provides a raster map compression system, comprising: a map acquisition module for acquiring an initial raster map; a map processing module for downsampling the initial raster map to obtain a compressed map, and encoding features of the initial raster map to obtain a first feature code; and a map update module for updating the initial raster map using the compressed map and the first feature code.

[0015] According to a third aspect of the present invention, the present invention further provides a vehicle comprising the above-mentioned grid map compression system.

[0016] According to the fourth aspect of the present invention, the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, it implements any one of the above-mentioned raster map compression methods.

[0017] According to another aspect of the present invention, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, any one of the above-mentioned raster map compression methods is implemented.

[0018] Through one or more of the above embodiments of the present invention, at least the following technical effects can be achieved:

[0019] The technical solution disclosed in the present invention can, on the one hand, downsample the raster map, thereby reducing the memory required for the raster map. On the other hand, the present invention encodes the initial raster map so that the features of each grid of the raster map are stored in an encoded form. When reading the compressed raster map, the features of the initial raster map stored in the encoded form can be read out, thereby avoiding the loss of accuracy caused by downsampling the raster map. Therefore, the present application can reduce the memory required for the raster map while ensuring that the accuracy of the raster map does not decrease. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The technical solutions and other beneficial effects of the present invention will be made apparent by describing in detail the specific embodiments of the present invention in conjunction with the accompanying drawings.

[0021] Figure 1 A flowchart of a raster map compression method provided by an embodiment of the present invention;

[0022] Figure 2 An example diagram of raster map downsampling in the raster map compression method provided by an embodiment of the present invention;

[0023] Figure 3 A schematic diagram of position encoding of a grid by a grid map compression method provided by an embodiment of the present invention;

[0024] Figure 4 A schematic diagram of a grid map compression method according to an embodiment of the present invention encoding a grid by probability values;

[0025] Figure 5 A schematic diagram of a grid map compression method according to an embodiment of the present invention using normal distribution to calculate occupancy probability values;

[0026] Figure 6 A schematic diagram of a grid map compression method according to an embodiment of the present invention using Gaussian distribution to calculate occupancy probability values;

[0027] Figure 7 This is a comparison chart of the 2D grid map before and after processing in the prior art;

[0028] Figure 8A comparison chart showing a 2D raster map before and after being processed by the raster map compression method provided in an embodiment of the present invention;

[0029] Figure 9 A schematic diagram of a 3D raster map processed by the raster map compression method provided in an embodiment of the present invention;

[0030] Figure 10 A framework diagram of a raster map compression system provided by an embodiment of the present invention;

[0031] Figure 11 This is a schematic block diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0033] In the description of the present invention, it should be noted that, unless otherwise specified or limited, the term "and / or" herein is merely a description of an association relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, and B exists alone. Furthermore, the character " / " herein, unless otherwise specified, generally indicates that the associated objects are in an "or" relationship.

[0034] Occupancy grid map, also called grid map, provides an effective method to express and process environmental information by converting spatial point coordinates into discrete grid form, so as to express information such as the boundaries and locations of obstacles.

[0035] In assisted and autonomous driving systems, perception modules that primarily focus on rectangular object detection cannot address long-tail problems such as detecting heterogeneous obstacles, obstacles of unknown types, and environmental obstacles. Occupancy grid maps, on the other hand, can better represent the boundaries between objects and obstacles. Consequently, occupancy grid maps are increasingly being used in point cloud perception and image perception systems.

[0036] Current autonomous driving perception systems have very high requirements for the range and accuracy of their surroundings. If grid maps are to meet these requirements, hardware requirements, especially memory requirements, are high. For some systems or products with limited hardware, the memory requirements of grid maps are the primary factor limiting their usefulness.

[0037] The amount of storage memory occupied by a raster map is mainly determined by the map size and the resolution of the raster. The following are examples of several typical point cloud systems:

[0038] system Length*Width*Height(m*m*m) Resolution (m) Memory size 360° millimeter-wave radar (2D) 200*200 0.5 625KB 360° LiDAR (3D) 200*200*20 0.2 381MB LiDAR SLAM map 1000*1000*20 0.2 9.31GB

[0039] As map sizes increase and resolution requirements rise, the memory requirements for occupancy grid maps will continue to grow. However, for some low-hardware perception systems without domain controllers or back-end processors, such as pure millimeter-wave radar perception systems, occupancy grid maps cannot be used.

[0040] Therefore, the embodiments of the present application provide a raster map compression method, system, vehicle, electronic device and storage medium, which can reduce the memory of the raster map while ensuring that the accuracy of the raster map does not decrease, thereby enabling the raster map to be used in some low-hardware perception systems without domain control or back-end processors.

[0041] Figure 1 The grid map compression method provided by an embodiment of the present invention includes:

[0042] S101, obtaining an initial grid map;

[0043] S102, downsampling the initial grid map to obtain a compressed map, and encoding features of the initial grid map to obtain a first feature code;

[0044] S103: Update the initial grid map using the compressed map and the first feature code.

[0045] In this embodiment, the initial grid map is obtained by initializing the original grid map.

[0046] In step S102, the initial raster map can be compressed by downsampling the initial raster map. However, if the raster map is only compressed, problems such as a decrease in the resolution of the raster map may occur. Therefore, while downsampling, this embodiment also encodes the features of the initial raster map, and uses the encoding to record the features of the initial raster map. In this way, no matter how the resolution of the compressed map decreases, the features of the initial raster map have been recorded by the first feature encoding. When using the raster map, the function of the original initial raster map can be realized by using the compressed map and the first feature encoding. Therefore, the memory space occupied by the raster map is greatly reduced by downsampling. Although the encoding is increased, the memory space occupied by the encoding is small, so the space of the raster map can be greatly reduced overall.

[0047] Therefore, the grid map compression method provided in this embodiment can, on the one hand, downsample the grid map, thereby reducing the memory required for the grid map. On the other hand, the present invention encodes the initial grid map so that the features of each grid of the grid map are stored in an encoded form. When reading the compressed grid map, the features of the initial grid map stored in the encoded form can be read out, thereby avoiding the loss of accuracy caused by downsampling the grid map. Therefore, this application can both reduce the memory required for the grid map and ensure that the accuracy of the grid map is not reduced.

[0048] In one embodiment, the step of downsampling the initial grid map includes:

[0049] Sampling the initial grid map using a first preset sampling window, wherein the first preset sampling window has a first predetermined number of grids;

[0050] According to the order in which the initial grid map is sampled, the map sampled each time in the first preset sampling window is projected into an area respectively, wherein the area corresponding to the projection of the map sampled in each first preset sampling window includes a second predetermined number of grids, and the first predetermined number is greater than the second predetermined number.

[0051] See also Figure 2 In this embodiment, the first predetermined number is 16, the second predetermined number is 4, and the first predetermined sampling window is a 4×4 sampling window. A single sampling window can capture 16 grids of the grid map. When the grid map is projected into an area, the 16 4×4 grids are projected into 4 grids, thereby downsampling the grid map and performing lossy compression on the grid map.

[0052] In other embodiments, the first preset number can also be 9, the second preset number is 1, the first preset sampling window is 3×3, or the first preset number can also be 16, the second preset number is 1, the first preset sampling window is 4×4, and the projection is 1, etc. It can be understood that the first preset number and the second preset number are a preset value, for example, the first number is any value between 2-64 or 65-128, etc., and the second predetermined number is any value less than the first predetermined number. It is worth noting that the value is only given as an example. As long as the second predetermined number is less than the first predetermined number, the value is within the protection scope of this application.

[0053] In one embodiment, the step of encoding the features of the initial grid map includes: encoding the position of each grid occupied feature in the initial grid map, and encoding the probability value of each grid occupied feature.

[0054] In this embodiment, by encoding the position of each grid, it is possible to obtain whether each grid is occupied by the map. By encoding the probability value of each grid, it is possible to obtain the probability of each grid being occupied by the map.

[0055] In one embodiment, the step of encoding the position of the occupied features of each grid in the initial grid map includes:

[0056] Get the blank grid and non-blank grid in the grid map;

[0057] A first mark is performed on the blank grid, and a second mark is performed on the non-blank grid, wherein the first mark and the second mark are different.

[0058] See also Figure 3 In this embodiment, the first mark can be 1 and the second mark can be 0. When encoding the initial grid map, the position encoding of the occupied features of each grid in the initial grid map is performed in order to mark whether each grid is occupied by the map. For example, if a grid is occupied by the map, that is, a non-blank grid, the position encoding of the grid is counted as 1. If a grid map is not occupied by the map, that is, a blank grid, the position encoding of the grid is counted as 0.

[0059] In one embodiment, the step of encoding the probability value of each grid being occupied includes:

[0060] Calculate the probability value of the map occupying each grid in the grid; use the numerical value of the probability value as the code.

[0061] In this embodiment, when encoding the probability value, the probability of each grid being occupied by the map is encoded. The accuracy of probability storage is related to the number of encoding bits, and the accuracy is 2 -N , N is the number of encoding bits. In this embodiment, N is 4. In other embodiments, N can also take other values, such as 1, 2, 3, 5, 6, 7, etc. It is worth noting that this value is only given as an example, and other values ​​are also within the scope of protection of this application. For details, please refer to Figure 4 , Figure 4 An example of probabilistic coding is given.

[0062] In this embodiment, the grid corresponding to the probability value code greater than a predetermined value is set as the occupied grid. For example, if the predetermined value is 0.50, the grid corresponding to the probability value greater than 0.50 is set as the occupied grid.

[0063] In other embodiments, encoding the probability value of each grid being occupied may further include obtaining non-blank grids with a second marker; calculating the percentage of the map in each non-blank grid; and using the percentage value as the code. Since the above embodiment has already determined whether each grid is occupied by a map, the percentage of the map in the grid may also be used to represent the characteristics of the grid.

[0064] In one embodiment, after encoding the features of the initial grid map to obtain the first feature codes, the method further includes: storing the first feature codes in a predetermined number and in a predetermined format.

[0065] When storing, when the sampling window is 4×4, the 16-bit position code and N-dimensional Huffman code are converted into N uint16 formats for storage, that is:

[0066]

[0067] In one embodiment, the map updating method further comprises: acquiring in real time the position state of the target vehicle and environmental point cloud data obtained by radar detection of the target vehicle;

[0068] Mapping environmental point cloud data into a raster map;

[0069] Re-encode the grid mapped by the environmental point cloud data to obtain a second feature code;

[0070] The first feature code of the grid mapped with the environmental point cloud data is updated using the second feature code.

[0071] In this embodiment, by acquiring the target vehicle's posture state and the environmental point cloud data obtained by radar detection of the target vehicle in real time, these data can be subjected to second feature encoding. These feature codes are attached as features to the grid map, so that the real-time acquired data can be updated to the grid map in a timely manner.

[0072] In this embodiment, the step of re-encoding the grid mapped by the environmental point cloud data includes:

[0073] The range and azimuth data of the environmental point cloud data are used to calculate the occupancy probability of the grid occupied by the point cloud data.

[0074] In one embodiment, during the calculation of the occupancy probability, the occupancy probability is calculated using the normal distribution of the range and azimuth data of the environmental point cloud data, or the occupancy probability is calculated using the Gaussian distribution of the range and azimuth data of the environmental point cloud data.

[0075] See also Figure 5When a vehicle or other carrier is in the grid map, the grid where the vehicle or carrier appears is used to update the needle probability corresponding to the grid. Specifically, when the normal distribution of the range and azimuth angle data of the environmental point cloud data is used to calculate the occupancy probability, the following formula can be used for calculation:

[0076]

[0077] Where x is the variable, μ is the mean of the probability density function, and σ is the standard deviation.

[0078] See also Figure 6 When a vehicle or other carrier is in the grid map, the grid where the vehicle or carrier appears is used to update the corresponding needle probability of the grid. Specifically, when the range of the environmental point cloud data and the Gaussian distribution of the azimuth angle data are used to calculate the occupancy probability, the following formula can be used for calculation:

[0079]

[0080] Among them, σ range is the range standard deviation in the azimuth data, σ zaimuth is the azimuth standard deviation in the azimuth data, P free is the constant value of the idle occupancy probability, P t is the occupation probability, P range is the Gaussian distribution of the range of the environmental point cloud data, P orth_range is the Gaussian distribution of azimuth data.

[0081] In one embodiment, the present application also tests and compares the above-mentioned grid map compression method, processes a 2D grid map in a millimeter wave radar system, and uses an initial grid map of 200 meters * 200 meters with a resolution of 0.25 meters. The initial grid map occupies 2500KB of memory. After the initial grid map is optimized by the above-mentioned grid map compression method, the memory occupied by the initial grid map is 312.5KB.

[0082] Specifically, see Figure 7 , Figure 7 represents a road intersection scene, and Figure 7 (Left) is the scene of the initial grid map. Figure 7 (Right) shows the grid map after optimization by the method in the prior art. It can be clearly seen that the resolution of the prior art method deteriorates after compressing the 2D grid map.

[0083] See also Figure 8 , Figure 8 represents a parking lot scene, and Figure 8 (Left) is the scene of the initial grid map. Figure 8(Right) shows the raster map after the raster map compression method of the present application is optimized. It can also be clearly seen that the resolution of the raster map compression method of the present application does not change significantly after compressing the 2D raster map, and vehicles in the parking lot can be easily identified.

[0084] In one embodiment, the present application also processes the 3D grid map in the millimeter wave radar map, using an initial grid map of 200 meters * 200 meters and a resolution of 0.5 meters. The grid map optimized by the grid map compression method of the present application occupies 3.05MB of memory. Figure 9 , it can be clearly seen from the figure that some features in the 3D image correspond to the real scene image.

[0085] Therefore, the grid map compression method provided in the embodiment of the present application can, on the one hand, downsample the grid map, thereby reducing the memory required for the grid map. On the other hand, the present invention encodes the initial grid map so that the features of each grid of the grid map are stored in encoded form. When reading the compressed grid map, the features of the initial grid map stored in the encoded form can be read out, thereby avoiding the loss of accuracy caused by downsampling the grid map. Therefore, the present application can both reduce the memory required for the grid map and ensure that the accuracy of the grid map is not reduced.

[0086] See also Figure 10 The embodiment of the present application also provides a raster map compression system, which includes a map acquisition module 1, a map processing module 2 and a map update module 3. The map acquisition module 1 is used to obtain an initial raster map; the map processing module 2 is used to downsample the initial raster map to obtain a compressed map, and encode the features of the initial raster map to obtain a first feature code; the map update module 3 is used to update the initial raster map using the compressed map and the first feature code.

[0087] In one embodiment, the map processing module includes: a downsampling unit and a projection unit; the downsampling unit is used to sample the initial grid map using a first preset sampling window, wherein the first preset sampling window has a first predetermined number of grids; the projection unit is used to project the map sampled each time by the first preset sampling window into an area in the order of sampling the initial grid map, wherein the area corresponding to the projection of the map sampled by each first preset sampling window includes a second predetermined number of grids, and the first predetermined number is greater than the second predetermined number.

[0088] In one embodiment, the map processing module further includes: a position encoding unit and a probability encoding unit, wherein the position encoding unit is used to perform position encoding on the occupied features of each grid in the initial grid map, and the probability encoding unit is used to encode the probability value of the occupied features of each grid.

[0089] In one embodiment, the position encoding unit includes: a grid acquisition subunit and a marking subunit; the grid acquisition subunit is used to acquire blank grids and non-blank grids in the grid map; the marking subunit is used to perform a first mark on the blank grid and a second mark on the non-blank grid, wherein the first mark and the second mark are different.

[0090] In one embodiment, the probability coding unit includes: a probability value calculation subunit and a probability value coding subunit; the probability value calculation subunit is used to calculate the probability value of the map occupying each grid; the probability value coding subunit is used to use the numerical value of the probability value as a code.

[0091] In one embodiment, the raster map compression system further includes: a storage module for storing the first feature codes in a predetermined number and in a predetermined format after the map processing module 2 encodes the features of the initial raster map to obtain the first feature codes.

[0092] In one embodiment, the raster map compression system further includes: a data acquisition module, a mapping module, a second feature encoding module and an update module, wherein the data acquisition module is used to acquire the posture state of the target vehicle and the environmental point cloud data obtained by radar detection of the target vehicle in real time; the mapping module is used to map the environmental point cloud data into the raster map; the second feature encoding module is used to re-encode the raster mapped with the environmental point cloud data to obtain a second feature code; and the update module is used to update the first feature code of the raster mapped with the environmental point cloud data using the second feature code.

[0093] In one embodiment, the second feature encoding module is specifically configured to calculate the occupancy probability of a grid occupied by the point cloud data using the range and azimuth data of the environmental point cloud data. The occupancy probability calculation process utilizes a normal distribution of the range and azimuth data of the environmental point cloud data, or utilizes a Gaussian distribution of the range and azimuth data of the environmental point cloud data.

[0094] The present application provides an electronic device. Figure 11 The electronic device includes: a memory 601, a processor 602, and a computer program stored in the memory 601 and executable on the processor 602. When the processor 602 executes the computer program, the raster map compression method described above is implemented.

[0095] Furthermore, the electronic device also includes: at least one input device 603 and at least one output device 604 .

[0096] The memory 601 , processor 602 , input device 603 , and output device 604 are connected via a bus 605 .

[0097] The input device 603 may be a camera, a touch panel, a physical button, a mouse, etc. The output device 604 may be a display screen.

[0098] The memory 601 may be a high-speed random access memory (RAM) or a non-volatile memory such as a disk drive. The memory 601 is used to store a set of executable program codes. The processor 602 is coupled to the memory 601.

[0099] Furthermore, embodiments of the present application also provide a computer-readable storage medium, which may be provided in the electronic device described in each of the above embodiments. The computer-readable storage medium may be the memory 601 described in the above embodiments. The computer-readable storage medium stores a computer program, which, when executed by the processor 602, implements the raster map compression method described in the above method embodiments.

[0100] Furthermore, the computer storable medium may also be various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory 601 (ROM), a RAM, a magnetic disk, or an optical disk.

[0101] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is merely a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0102] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of the present embodiment according to actual needs.

[0103] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or software functional modules.

[0104] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in various embodiments of the present invention.

[0105] It should be noted that for the aforementioned method embodiments, for ease of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0106] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0107] In summary, although the present invention has been disclosed above with reference to preferred embodiments, the above preferred embodiments are not intended to limit the present invention. A person skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined in the claims.

Claims

1. A raster map compression method, characterized in that: include: Get the initial grid map; Downsampling the initial grid map to obtain a compressed map, and encoding features of the initial grid map to obtain a first feature code; The initial grid map is updated using the compressed map and the first feature code.

2. The raster map compression method according to claim 1, wherein: The step of downsampling the initial grid map comprises: Sampling the initial grid map using a first preset sampling window, wherein the first preset sampling window has a first predetermined number of grids; According to the order in which the initial grid map is sampled, the map sampled each time by the first preset sampling window is projected into an area respectively, wherein the area corresponding to the projection of the map sampled by each first preset sampling window includes a second predetermined number of grids, and the first predetermined number is greater than the second predetermined number.

3. The raster map compression method according to claim 1, wherein: The step of encoding the features of the initial grid map comprises: The position of each grid occupied feature in the initial grid map is encoded, and the probability value of each grid occupied feature is encoded.

4. The raster map compression method according to claim 3, wherein: The step of encoding the position of the occupied features of each grid in the initial grid map comprises: Obtain blank grids and non-blank grids in the grid map; The blank grid is marked with a first mark, and the non-blank grid is marked with a second mark, wherein the first mark and the second mark are different.

5. The raster map compression method according to claim 3, wherein: The step of encoding the probability value of each grid being occupied comprises: Calculate the probability value of the map occupying each grid; The numerical value of the probability value is encoded as the probability value of the occupancy feature.

6. The raster map compression method according to claim 1, wherein: After encoding the features of the initial grid map to obtain the first feature codes, the method further includes: storing the first feature codes in a predetermined number and in a predetermined format.

7. The raster map compression method according to claim 1, wherein: The method further comprises: acquiring in real time the position state of the target vehicle and environmental point cloud data obtained by radar detection of the target vehicle; Mapping the environmental point cloud data into the grid map; Re-encoding the grid mapped by the environmental point cloud data to obtain a second feature code; The first feature code of the grid mapped with the environmental point cloud data is updated using the second feature code.

8. The raster map compression method according to claim 7, wherein: The step of re-encoding the grid mapped by the environmental point cloud data comprises: The range and azimuth data of the environmental point cloud data are used to calculate the occupancy probability of the grid occupied by the point cloud data.

9. The raster map compression method according to claim 8, wherein: During the calculation of the occupancy probability, the occupancy probability is calculated using the normal distribution of the range and azimuth data of the environmental point cloud data, or the occupancy probability is calculated using the Gaussian distribution of the range and azimuth data of the environmental point cloud data.

10. A raster map compression system, characterized in that: include: Map acquisition module, used to obtain the initial raster map; a map processing module, configured to downsample the initial grid map to obtain a compressed map, and encode features of the initial grid map to obtain a first feature code; A map updating module is configured to update the initial grid map using the compressed map and the first feature code.

11. A vehicle, characterized in that: Comprising the raster map compression system as claimed in claim 10.

12. An electronic device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images