Method and device for determining drivable area, storage medium and electronic equipment

Abstract

The invention discloses a method and device for determining a drivable area, a storage medium and electronic equipment. The method comprises the following steps: determining a first attention area of a vehicle at a first moment, an occupation probability of each grid in a first occupation grid map corresponding to the first attention area, and driving data of the vehicle at a second moment after the first moment; determining a second region of interest of the vehicle at a second moment according to the first region of interest and the driving data; determining the occupancy probability of each grid in a second occupancy grid map corresponding to the second region of interest according to the driving data and the occupancy probability of each grid in the first occupancy grid map; and according to the occupancy probability of each grid in the second occupancy grid map, determining a drivable area in the second concern area. According to the method, feature fusion does not need to be carried out on different-source driving data, the data calculation amount can be reduced, the driving area can be determined with low consumption and low time consumption, and the problems of high fusion difficulty and long time consumption when feature fusion is carried out on different-source driving data are solved.

Description

Technical Field

[0001] The present disclosure relates to the field of intelligent driving technology, and in particular to a method, device, storage medium and electronic device for determining a drivable area. Background Art

[0002] Smart driving vehicles usually use a variety of sensors such as on-board cameras, lidar, millimeter-wave radar, ultrasonic radar, etc. to collect data about the vehicle's surrounding environment. The smart driving system performs environmental perception, analysis and decision-making, and intelligent control based on the environmental data collected by the vehicle's various sensors to achieve automatic or semi-automatic driving of the vehicle. Since the data collected by sensors from different sources have differences in type and dimension, they need to be fused when determining the drivable area. In the related technology, when fusing data of different types and dimensions collected by sensors from different sources, there are defects such as large consumption of computing resources, long time consumption, and poor fusion effect, which results in the determination of the drivable area being not accurate enough, affecting the safe and stable driving of smart driving vehicles. Summary of the invention

[0003] In order to solve the technical problems of high resource consumption, long time consumption, poor fusion effect, etc. in the related art when fusing sensor data from different sources, the present invention provides a method, device, storage medium and electronic device for determining a drivable area, which can realize low-consumption fusion processing of multi-sensor data, improve the fusion effect, and help improve the accuracy of the drivable area.

[0004] An embodiment of the first aspect of the present disclosure provides a method for determining a drivable area, including:

[0005] Determine a first focus area of ​​the vehicle at a first moment, an occupancy probability of each grid in a first occupancy grid map corresponding to the first focus area, and driving data of the vehicle at a second moment; wherein the second moment is later than the first moment;

[0006] determining a second area of ​​interest for the vehicle at the second moment according to the first area of ​​interest and the driving data;

[0007] Determining, according to the driving data and the occupancy probability of each grid in the first occupancy grid map, the occupancy probability of each grid in the second occupancy grid map corresponding to the second area of ​​interest;

[0008] A drivable area within the second area of ​​interest is determined according to the occupancy probability of each grid in the second occupancy grid map.

[0009] An embodiment of a second aspect of the present disclosure provides a device for determining a drivable area, including:

[0010] A first determination module is used to determine a first focus area of ​​the vehicle at a first moment, an occupancy probability of each grid in a first occupancy grid map corresponding to the first focus area, and driving data of the vehicle at a second moment; wherein the second moment is later than the first moment;

[0011] A second determining module, configured to determine a second area of ​​interest of the vehicle at the second moment according to the first area of ​​interest and the driving data;

[0012] a third determination module, configured to determine an occupancy probability of each grid in a second occupancy grid map corresponding to the second area of ​​interest based on the driving data and the occupancy probability of each grid in the first occupancy grid map;

[0013] The fourth determination module is used to determine the drivable area in the second area of ​​interest according to the occupancy probability of each grid in the second occupancy grid map.

[0014] An embodiment of the third aspect of the present disclosure provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to execute the method for determining a drivable area proposed in the embodiment of the first aspect of the present disclosure.

[0015] In a fourth aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0016] processor;

[0017] a memory for storing instructions executable by the processor;

[0018] The processor is used to read the instruction from the memory and execute the instruction to implement the method for determining the drivable area proposed in the embodiment of the first aspect of the present disclosure.

[0019] A fifth aspect embodiment of the present disclosure provides a computer program product, which, when an instruction processor in the computer program product executes, executes the method for determining a drivable area proposed in the first aspect embodiment of the present disclosure.

[0020] In the disclosed embodiment, a device for determining a drivable area first determines a first area of ​​interest of the vehicle at a first moment, an occupation probability of each grid in a first occupation grid map corresponding to the first area of ​​interest, and driving data of the vehicle at a second moment after the first moment; then determines a second area of ​​interest of the vehicle at the second moment based on the first area of ​​interest and the driving data; then determines an occupation probability of each grid in a second occupation grid map corresponding to the second area of ​​interest based on the driving data and the occupation probability of each grid in the first occupation grid map; and finally determines a drivable area within the second area of ​​interest based on the occupation probability of each grid in the second occupation grid map. The scheme determines the probability of grid occupation based on the driving data in the focus area collected by different source sensors at each moment, and performs probability fusion based on the probability of grid occupation to obtain the occupation probability of each grid, and then determines the drivable area. Through the preset focus area, only the driving data in the focus area at the current moment needs to be considered at each moment, which can reduce the amount of data calculation and achieve low-consumption and low-time determination of the drivable area. Moreover, the scheme fuses the occupation probability when determining the occupation probability of each grid in the second occupied grid map, and the fusion difficulty is small. There is no need to perform feature fusion processing on data of different types and dimensions, which can solve the problem of difficulty in fusion due to differences in data types and dimensions when performing feature fusion on driving data from different sources. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a schematic diagram of the composition architecture of a system for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0022] Figure 2 It is a schematic diagram of constructing an occupancy grid map by a device for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0023] Figure 3 is a flowchart of a method for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0024] Figure 4 is a flowchart of a method for determining a drivable area provided by another exemplary embodiment of the present disclosure;

[0025] Figure 5 is a schematic diagram of a first region of interest and a second region of interest provided by an exemplary embodiment of the present disclosure;

[0026] Figure 6 is a flowchart of a method for determining a drivable area provided by yet another exemplary embodiment of the present disclosure;

[0027] Figure 7 It is a schematic diagram of determining the update probability of a grid according to each source data in a method for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0028] Figure 8 is a flowchart of a method for determining a drivable area provided by another exemplary embodiment of the present disclosure;

[0029] Fig. 9 is a schematic diagram of grid coordinates of each characteristic contour point in a grid map coordinate system provided by an exemplary embodiment of the present disclosure;

[0030] Fig.10 is a schematic diagram of an interpolation principle in a method for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0031] Fig.11 is a schematic diagram of grid coordinates of characteristic target contour points obtained by interpolation in a method for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0032] Fig.12 It is a schematic diagram of storing data in a compressed manner in a method for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0033] Fig.13 is a flowchart of a method for determining a drivable area provided by yet another exemplary embodiment of the present disclosure;

[0034] Fig.14 is a flowchart of a method for determining a drivable area provided by another exemplary embodiment of the present disclosure;

[0035] Fig.15 is a schematic diagram of a scene for resampling in a method for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0036] Fig.16 is a schematic diagram of the composition structure of a device for determining a drivable area provided by an exemplary embodiment of the present disclosure;

[0037] Fig.17 is a schematic diagram of the composition structure of a device for determining a drivable area provided by another exemplary embodiment of the present disclosure;

[0038] Fig.18 It is a schematic diagram of the composition structure of an electronic device provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0039] To explain the present disclosure, example embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. It should be understood that the present disclosure is not limited to the example embodiments.

[0040] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0041] Application Overview

[0042] The intelligent driving system is an assisted driving system developed based on automatic control technology. It makes driving decisions autonomously by sensing and analyzing the environmental information around the intelligent driving vehicle (hereinafter referred to as the vehicle), thereby realizing automatic or semi-automatic driving of the vehicle.

[0043] In order to accurately and comprehensively perceive the environmental information around the vehicle during driving, various sensors such as vehicle-mounted cameras, laser radars, millimeter-wave radars, and ultrasonic radars are generally used to collect data about the vehicle's surrounding environment. The types and dimensions of the various data collected by different types of sensors are different. For example, the types of images collected by vehicle-mounted cameras and echo signals collected by ultrasonic radars are different. For example, the dimensions of two-dimensional images collected by vehicle-mounted cameras and three-dimensional point clouds collected by millimeter-wave radars are different.

[0044] When determining the drivable area, the intelligent driving system needs to fuse and process multi-source data of different types and dimensions collected by various sensors. The commonly used fusion solutions in related technologies mainly include the following two:

[0045] One fusion solution is to convert the data collected by each sensor into structured information, such as a bounding box in a two-dimensional coordinate system, which describes the position, size, shape and other information of the object. Then, a filter is used to fuse the various structured information.

[0046] This fusion scheme converts the extracted target objects into regular bounding boxes, so the target objects to be extracted are all relatively standard geometric structures. For obstacles such as vehicles and pedestrians that are on the whitelist, they can be structured using geometric structures such as rectangular boxes. However, for obstacles outside the whitelist (such as large plastic bags on the road), it is difficult to structure them using general geometric structures in most cases. This fusion scheme is not suitable for obstacles with non-standard structures. In addition, the acquisition frequencies of different sensors are generally different. Before the filter is fused, the frequencies corresponding to each structured information need to be aligned, otherwise the accuracy of the fusion result will be affected.

[0047] Another fusion scheme is to convert the data collected by different sensors into an intermediate form, such as extracting features from different source data to obtain features corresponding to different source data, then converting features corresponding to different source data to obtain conversion features, and then fusing the conversion features to obtain fusion features, such as fusion feature point sets or fusion feature layers, and then determining the occupancy probability of each grid in the grid map based on the fusion features to obtain the current round of occupancy probability, and finally determining the occupancy probability by superimposing the grid map probabilities. This fusion method consumes too much computing resources and takes a long time, and is not suitable for scenarios with high real-time requirements.

[0048] In order to solve the technical problems existing in the related art that the applicable scenarios are limited and data of different types and dimensions collected by multi-source sensors cannot be integrated with low consumption, low time consumption and accuracy, which leads to the inability to accurately determine the drivable area and affects the safe and stable driving of the vehicle, a method for determining the drivable area is provided in an embodiment of the present disclosure. First, a first focus area of ​​the vehicle at a first moment, the occupation probability of each grid in a first occupation grid map corresponding to the first focus area, and the driving data of the vehicle at a second moment after the first moment are determined; then, based on the first focus area and the driving data, a second focus area of ​​the vehicle at the second moment is determined; then, based on the driving data and the occupation probability of each grid in the first occupation grid map, the occupation probability of each grid in the second occupation grid map corresponding to the second focus area is determined; finally, based on the occupation probability of each grid in the second occupation grid map, the drivable area within the second focus area is determined. The method determines the probability of grid occupancy based on the driving data in the focus area collected by different source sensors at each moment, performs probability fusion based on the probability of grid occupancy to obtain the occupancy probability of each grid, and then determines the drivable area. Through the preset focus area, only the driving data in the focus area at the current moment needs to be considered at each moment, which can reduce the amount of data calculation and achieve low-consumption and low-time determination of the drivable area. In addition, the scheme fuses the occupancy probability, with low fusion difficulty. There is no need to perform feature fusion processing on data of different types and dimensions, which can solve the problem of difficulty in fusion due to differences in data types and dimensions when performing feature fusion on driving data from different sources.

[0049] Exemplary Systems

[0050] Figure 1 FIG. 1 is a schematic diagram of the composition architecture of a system for determining a drivable area provided by an exemplary embodiment of the present disclosure. Figure 1As shown, the system 10 for determining a drivable area includes a data acquisition device 100 and a drivable area determination device 200. The data acquisition device 100 and the drivable area determination device 200 can be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiments of the present disclosure. The data acquisition device 100 is used to collect driving data, and the drivable area determination device 200 is used to determine the drivable area based on the driving data collected by the data acquisition device 100.

[0051] For example, Figure 1 As shown, the data acquisition device 100 and the drivable area determination device 200 can be set on the intelligent driving vehicle (referred to as the vehicle), and data transmission is achieved through a wired link. Alternatively, in other embodiments, the data acquisition device 100 is set on the vehicle, and the drivable area determination device 200 can be set outside the vehicle, such as on the server side, and the data acquisition device 100 and the drivable area determination device 200 achieve data transmission through a wireless link.

[0052] The data collection device 100 includes a plurality of different types of sensors for collecting different types of vehicle driving data. For example, Figure 1 As shown, the data acquisition device 100 may include sensors such as a vehicle-mounted camera 101, a millimeter-wave radar 102, a laser radar 103, and an ultrasonic radar 104. During the driving process of the vehicle, the vehicle-mounted camera 101, the millimeter-wave radar 102, the laser radar 103, and the ultrasonic radar 104 collect data of the vehicle's surrounding environment in real time, and the data acquisition device 100 obtains driving data from different sources. The data acquisition device 100 transmits the driving data from different sources collected in real time to the driving area determination device 200.

[0053] The drivable area determination device 200 receives the driving data transmitted by the data acquisition device 100 in real time, and determines the drivable area of ​​the vehicle at the current moment according to the received driving data. Before describing the drivable area determination step, the way in which the drivable area determination device 200 manages data is described first.

[0054] Since the vehicle usually only pays attention to the obstacle information within a limited range around it during driving, based on this, a focus area is set in the embodiment of the present disclosure, and the area around the vehicle can be divided into multiple grids to construct a grid map corresponding to the focus area. At each moment, it is only necessary to determine the occupancy of each grid in the grid map corresponding to the focus area, and there is no need to determine the occupancy of the grids outside the focus area. The drivable area is then determined based on the occupancy of the grids in the focus area, which can reduce redundant data and reduce the amount of data calculation. Figure 2 It is a schematic diagram of constructing an occupancy grid map by a device for determining a drivable area provided by an exemplary embodiment of the present disclosure.

[0055] When the device for determining the drivable area constructs an occupancy grid map, it first constructs an initial occupancy grid map; then, whenever driving data transmitted by the data acquisition device is received, the occupancy grid map constructed last time is updated according to the driving data received in real time, so as to determine the occupancy probability of each grid in the occupancy grid map corresponding to each moment in the vehicle driving process, and then determine the drivable area of ​​the vehicle at each moment according to the occupancy probability of each grid.

[0056] Exemplarily, before the vehicle travels, when the device for determining the drivable area constructs the initial occupancy grid map, it first determines the current position of the vehicle, wherein the current position of the vehicle may be the coordinates of the vehicle in the world coordinate system, also referred to as vehicle coordinates, and specifically may be the coordinates of the center point of the rear axle of the vehicle in the world coordinate system. Then, the area of ​​interest is determined according to the current position of the vehicle. The area of ​​interest is the surrounding area that the vehicle is concerned about at the current moment and is related to the passage of the vehicle. Exemplarily, the current position of the vehicle and the area of ​​interest may be as follows: Figure 2 As shown in (a) in FIG. 1 , the size of the focus area can be a preset size.

[0057] Exemplarily, the occupancy grid map corresponding to the area of ​​interest can be constructed using a grid map based on tile technology. The side length of the grid is determined according to the actual scenario, that is, the accuracy of dividing the occupied grid map, and then the entire area of ​​interest is mapped to the occupied grid map to obtain a grid map divided into multiple equal-length sides. The higher the grid division accuracy (that is, the smaller the grid side length), the higher the resolution of the map, but the amount of calculation will also increase accordingly. Therefore, in practical applications, it is necessary to weigh the relationship between the resolution of the map and the amount of calculation to select a suitable grid division accuracy.

[0058] The initial occupancy grid map constructed is as follows Figure 2 As shown in (b), the upper left corner point of the grid map is the origin O of the grid map coordinate system, the positive direction of the horizontal axis MX of the grid map coordinate system is toward the right side of the vehicle, and the positive direction of the vertical axis MY is toward the opposite direction of the vehicle's travel direction. The mapping grid of the center point of the rear axle of the vehicle in the initial occupied grid map is grid 201. Exemplarily, the occupied grid map can be divided into multiple sub-graphs. The device for determining the drivable area determines the sub-regions corresponding to each sub-graph and the world coordinates of the origin of the occupied grid map. Among them, the length and width of the sub-regions corresponding to each sub-graph may be completely the same, or may be partially the same (such as the same length or the same width), and the sum of the areas of the sub-regions corresponding to all sub-graphs is the area of ​​interest. Exemplarily, Figure 2 (b) in FIG. 1 is an example of an occupied grid map with a nine-grid layout and sub-regions of the same size corresponding to each sub-image. The world coordinates (worldx, worldy) of the origin O are the world coordinates of the origin of the occupied grid map. For example, Figure 2In (b), the origin of the grid map is shown as the upper left corner point, and its world coordinates are the world coordinates of the upper left corner point of the focus area, which can be determined according to the vehicle coordinates and the size of the focus area. Alternatively, the sub-graphs divided by the grid map can be a nine-grid layout, or a 4-grid, 16-grid, 25-grid or other layout, which is not limited in the embodiment of the present disclosure.

[0059] The device for determining the drivable area allocates a storage space for each sub-graph to store the data of each sub-graph. The storage space of each sub-graph may be a continuous storage space or a discontinuous storage space. For example, Figure 2 As shown in (b), the sub-images initially occupying the grid map are numbered 0 to 8 in order of row and column. Figure 8 The allocated storage spaces are storage space 0 to storage space 8 respectively.

[0060] During vehicle driving, the area of ​​interest will change with the vehicle's position. After receiving the driving data collected in real time by the data acquisition device, the device for determining the drivable area determines whether the vehicle's area of ​​interest has changed based on the currently received driving data. If it has changed, the sub-image combination order of the grid map needs to be updated. Exemplarily, the device for determining the drivable area can determine whether the area of ​​interest has changed based on whether the center point of the rear axle of the vehicle exceeds the movable range corresponding to the area of ​​interest at the previous moment. Among them, the movable range is the basis for determining whether the area of ​​interest needs to be updated, and its corresponding boundary line in the grid map is as follows: Figure 2 If the mapping grid of the center point of the rear axle of the vehicle in the occupied grid map is within the dotted box, that is, the center point of the rear axle of the vehicle is within the movable range, it is considered that the vehicle's area of ​​interest has not changed, and there is no need to update the vehicle's area of ​​interest, so there is no need to update the sub-graph combination order of the occupied grid map; if the mapping grid of the center point of the rear axle of the vehicle in the occupied grid map is outside the dotted box, that is, the center point of the rear axle of the vehicle is outside the movable range, it is considered that the vehicle's area of ​​interest has changed, and the vehicle's area of ​​interest needs to be updated, so the sub-graph combination order of the occupied grid map needs to be updated.

[0061] For example, when the vehicle is moving to the right, the center point of the rear axle moves to the Figure 5 The vehicle is still in the corresponding sub-area and is still within the movable range. The vehicle's focus area has not changed, and the occupancy grid map corresponding to the focus area remains unchanged. At this time, the mapping grid of the vehicle coordinates in the occupancy grid map is as follows Figure 2For the grid 202 shown in (c) in the figure, the device for determining the drivable area only needs to update the occupancy probability of each grid in the occupied grid map. Specifically, the device for determining the drivable area determines the occupancy probability of each grid in the current occupied grid map according to the driving data collected by the different source sensors currently received; then fuses the occupancy probabilities corresponding to the data of the different source sensors to obtain the updated probability; finally, updates the occupancy probability of each grid in the current occupied grid map according to the updated probability to obtain the new occupancy probability of each grid in the current occupied grid map. At this point, the map is updated once, and a new occupied grid map is obtained.

[0062] exist Figure 2 Based on (c) in Figure 5 Continue to drive right in the corresponding sub-area. When the center point of the rear axle of the vehicle moves to the sub-area Figure 5 When the vehicle is in the corresponding sub-area and outside the movable range, the vehicle's focus area changes, and the occupancy grid map corresponding to the focus area changes. At this time, the device for determining the drivable area not only needs to update the occupancy grid map, but also needs to update the occupancy probability of each grid in the updated occupancy grid map. Specifically, the device for determining the drivable area updates the world coordinates of the origin of the occupancy grid map according to the current position of the vehicle, and moves rightward from the upper left corner of the previous sub-map 0 to the sub-map 1. Figure 1 The upper left corner of the corresponding movable range is shifted to the right by an equal length, such as Figure 2 As shown in (c) and (d) in Figure 2 (c) in Figure 4 The position when it is the center, move to Figure 2 (d) in Figure 5 The first column of subgraphs (subgraph 0, subgraph 1, Figure 3 Kazuko Figure 6 ), spliced ​​to the last column of subgraphs (sub Figure 2 ,son Figure 5 Kazuko Figure 8 ) to get the updated Figure 2 The order of rows and columns shown in (d) is Figure 1 ,son Figure 2 , sub-image 0, sub- Figure 4 ,son Figure 5 ,son Figure 3 ,son Figure 7 ,son Figure 8 Kazuko Figure 6The grid map of the vehicle coordinates in the grid map is grid 203. After the order of the rows and columns of each sub-graph in the grid map is updated, the storage space is updated synchronously, and the storage space of the original sub-graph that does not need to be maintained is allocated to the new sub-graph that needs to be maintained. Figure 3 Kazuko Figure 6 The corresponding storage space is used to store the new sub-graph 0 and sub-graph Figure 3 Kazuko Figure 6 Then, according to the currently received driving data, the update probability of each grid in the new occupied grid map is determined, and the occupation probability of each grid in the new occupied grid map is updated according to the updated probability to obtain the occupation probability of each grid in the new occupied grid map. At this point, another update of the map is completed, and a new occupied grid map is obtained.

[0063] exist Figure 2 Based on (c) in Figure 5 Continue to drive upward to the right in the corresponding sub-area. When the center point of the rear axle of the vehicle moves to the sub-area Figure 2 When the vehicle is in the corresponding sub-area and outside the movable range, the vehicle's focus area changes, and the occupancy grid map corresponding to the focus area changes. At this time, the device for determining the drivable area not only needs to update the occupancy grid map, but also needs to update the occupancy probability of each grid in the updated occupancy grid map. Specifically, the device for determining the drivable area updates the coordinates of the origin of the occupancy grid map according to the current position of the vehicle, and moves from the upper left corner of the previous sub-map 0 to the right to the sub-map 1. Figure 1 The upper left corner of the sub-image is then translated upward by the distance of a sub-image. The corresponding boundary line of the movable range is represented by the sub-image. Figure 4 When the center is moved to the sub Figure 2 When updating the grid map, you can first splice the first column of sub-images to the right of the last column of sub-images, and then splice the last row of sub-images to the top of the first row of sub-images; you can also first splice the last row of sub-images to the top of the first row of sub-images, and then splice the first column of sub-images to the right of the last column of sub-images, and get the updated Figure 2 The order of rows and columns shown in (e) is Figure 7 ,son Figure 8 ,son Figure 6 ,son Figure 1 ,son Figure 2 , sub-image 0, sub- Figure 4 ,son Figure 5 Kazuko Figure 3The grid map of the vehicle coordinates in the grid map is grid 204. After the order of the rows and columns of each sub-graph in the grid map is updated, the storage space is updated synchronously, and the storage space of the original sub-graph that does not need to be maintained is allocated to the new sub-graph that needs to be maintained. Figure 3 ,son Figure 6 ,son Figure 7 Kazuko Figure 8 The corresponding storage space is used to store the new sub-graph 0 and sub-graph Figure 3 ,son Figure 6 ,son Figure 7 Kazuko Figure 8 Then, according to the currently received driving data, the update probability of each grid in the new occupied grid map is determined, and the occupation probability of each grid in the new occupied grid map is updated according to the updated probability to obtain the occupation probability of each grid in the new occupied grid map. At this point, another update of the map is completed, and a new occupied grid map is obtained.

[0064] Based on the above updating principle, during the driving process of the vehicle, the device for determining the drivable area continuously updates the occupancy grid map based on the received driving data at each moment, and obtains the occupancy grid map corresponding to each moment.

[0065] It should be noted that the size of the area of ​​interest and the movable range can be defined according to different driving scenarios, safety levels, and computing power. For example, if the safety level requirement is high, the area of ​​interest is large and the movable range is small. Conversely, different definitions can be used for scenes such as highways and urban areas. This information can be adjusted based on the scene combined with experience. In each sub-map of the occupied grid map, the occupancy probability of each grid in the newly constructed sub-map can be set to the initial value. Exemplarily, the initial value of the occupancy probability can be 0.5, that is, when no information is observed, the occupancy probability and idle probability of each grid in the newly constructed sub-map are set to the initial value of 0.5, and the probability of each grid being occupied and not occupied is equal.

[0066] In the embodiment of the present disclosure, the device for determining the drivable area is based on Figure 2 The principle shown here constructs an occupancy grid map corresponding to the area of ​​interest at each moment, determines the occupancy probability of each grid in the occupancy grid map based on the driving data collected by different source sensors at each moment, and determines the drivable area based on the occupancy probability. By presetting the area of ​​interest, only the driving data within the area of ​​interest at the current moment needs to be considered at each moment, which can reduce the amount of data calculation; multiple sub-graphs of the occupancy grid map are used to maintain data, and the occupancy grid map is updated by recombining multiple sub-graphs without increasing storage space, so as to realize dynamic following of the drivable area and provide data support for subsequent trajectory planning based on the drivable area.

[0067] Exemplary Methods

[0068] Figure 3 is a flowchart of a method for determining a drivable area provided by an exemplary embodiment of the present disclosure. Figure 1 The drivable area is determined on the device shown, such as Figure 3 As shown, the method for determining the drivable area may include the following steps:

[0069] Step S301, determining a first focus area of ​​a vehicle at a first moment, an occupancy probability of each grid in a first occupancy grid map corresponding to the first focus area, and driving data of the vehicle at a second moment.

[0070] The embodiments of the present disclosure may be executed by a device for determining a drivable area. Exemplarily, the device for determining a drivable area may be disposed on a vehicle and receive driving data collected by a data collection device on the vehicle.

[0071] Exemplarily, the device for determining the drivable area can read from the memory the first area of ​​interest previously determined and the occupancy probability of each grid in the first occupancy grid map corresponding to the first area of ​​interest. The first area of ​​interest is the surrounding area that the vehicle is concerned about at the first moment and is related to the passage of the vehicle. The first area of ​​interest can be determined based on the position of the vehicle at the first moment. The map constructed by the device for determining the drivable area for the first area of ​​interest is called the first occupancy grid map. The occupancy probability of each grid in the first occupancy grid map represents the possibility of the existence of an obstacle in the spatial area corresponding to each grid in the first area of ​​interest. The occupancy probability of each grid in the first occupancy grid map can be determined based on the driving data of the vehicle at the first moment.

[0072] Exemplarily, the device for determining the drivable area can obtain the driving data of the vehicle at the second moment collected by the data acquisition device. Among them, the data acquisition device may include a variety of different types of sensors for collecting different types of vehicle driving data. Exemplarily, the data acquisition device may include sensors such as vehicle-mounted cameras, millimeter-wave radars, laser radars, and ultrasonic radars. The device for determining the drivable area obtains the data collected by each sensor of the data acquisition device at the second moment, and obtains the driving data of the vehicle at the second moment. Therefore, the driving data includes different source data collected by different sensors, including but not limited to at least one of the following: image data collected by the vehicle-mounted camera, millimeter-wave radar data collected by the millimeter-wave radar, laser radar data collected by the laser radar, ultrasonic radar data collected by the ultrasonic radar, etc.

[0073] The first moment is the moment before the second moment, that is, the second moment is later than the first moment. Exemplarily, the interval between the second moment and the first moment can be the transmission period of the data acquisition device, and the transmission period can be equal to the acquisition period, and the data is transmitted once each time it is acquired; the transmission period can also be an integer multiple of the acquisition period, and the data is transmitted once after multiple acquisitions.

[0074] Step S302: determining a second area of ​​interest of the vehicle at a second moment according to the first area of ​​interest and the driving data.

[0075] Exemplarily, the data acquisition device may also include a position detector. In addition to the different source data collected by sensors such as the vehicle camera, millimeter wave radar, laser radar, and ultrasonic radar at the second moment, the corresponding driving data may also include the position data at the second moment obtained by the position detector detecting the position of the vehicle. The device for determining the drivable area may determine the second area of ​​interest of the vehicle at the second moment based on the first area of ​​interest of the vehicle at the first moment and the position of the vehicle at the second moment. The second area of ​​interest is the surrounding area that the vehicle is concerned about at the second moment and is related to the passage of the vehicle.

[0076] Step S303: determining the occupancy probability of each grid in the second occupancy grid map corresponding to the second area of ​​interest according to the driving data and the occupancy probability of each grid in the first occupancy grid map.

[0077] Among them, the occupancy probability of each grid in the second occupancy grid map represents the possibility of the existence of an obstacle in the spatial area corresponding to each grid in the second focus area, which can be determined based on the update probability determined by the driving data at the second moment and the occupancy probability of each grid in the first occupancy grid map determined before the second moment.

[0078] When determining the occupancy probability of each grid in the second occupied grid map, the device for determining the drivable area first determines the occupancy of each grid corresponding to each source data according to the driving data collected by different source sensors, for example, the first probability of each grid being occupied is determined according to the image data, the second probability of each grid being occupied is determined according to the millimeter wave radar data, the third probability of each grid being occupied is determined according to the laser radar data, and the fourth probability of each grid being occupied is determined according to the ultrasonic radar data; then the first probability, the second probability, the third probability and the fourth probability of each grid corresponding to each source data are fused to obtain the updated probability of each grid corresponding to the driving data at the second moment. Finally, the occupancy probability of each grid in the second occupied grid map corresponding to the second area of ​​interest is determined according to the updated probability of each grid and the occupancy probability of each grid in the first occupied grid map. In the embodiment of the present disclosure, the occupancy probability is fused when determining the occupancy probability, which has low fusion difficulty, low consumption and short time consumption. There is no need to perform feature fusion processing on data of different types and dimensions, which can solve the problem of high fusion difficulty due to differences in data types and dimensions when performing feature fusion on driving data of different sources.

[0079] Step S304: determining a drivable area within the second area of ​​interest according to the occupancy probability of each grid in the second occupancy grid map.

[0080] After determining the drivable area, the device determines the occupancy probability of each grid in the second occupancy grid map, and determines the state of each grid according to the occupancy probability of each grid. The state of the grid includes an occupied state and an unoccupied state. Target grids in an unoccupied state are selected from the grids in the second occupancy grid map, and the spatial areas corresponding to all target grids are formed into the area where the vehicle can pass at the second moment, so as to obtain the drivable area in the second area of ​​interest. For the target grid in an unoccupied state, there are no obstacles in the corresponding spatial area, and the vehicle is allowed to pass.

[0081] During the driving process of the vehicle, the driving data is continuously collected through the data collection device, and the device for determining the drivable area continuously updates the occupancy grid map according to the driving data at each moment, and then continuously determines the drivable area in each area of ​​interest. At each moment, only the driving data in the area of ​​interest at the current moment needs to be considered, which can reduce the amount of data calculation. Without increasing the storage space, the dynamic tracking of the drivable area can be achieved by updating the occupancy grid map, providing data support for subsequent trajectory planning based on the drivable area.

[0082] The method for determining a drivable area provided by an embodiment of the present disclosure first determines a first area of ​​interest of a vehicle at a first moment, an occupation probability of each grid in a first occupation grid map corresponding to the first area of ​​interest, and driving data of the vehicle at a second moment after the first moment; then, based on the first area of ​​interest and the driving data, determines a second area of ​​interest of the vehicle at the second moment; then, based on the driving data and the occupation probability of each grid in the first occupation grid map, determines the occupation probability of each grid in a second occupation grid map corresponding to the second area of ​​interest; finally, based on the occupation probability of each grid in the second occupation grid map, determines a drivable area within the second area of ​​interest. The method constructs an occupancy grid map corresponding to the focus area. At each moment, only the driving data in the focus area at the current moment needs to be considered, which can reduce the amount of data calculation. Without increasing the storage space, the method realizes dynamic following of the drivable area by updating the occupancy grid map, and provides data support for subsequent trajectory planning based on the drivable area. In addition, when determining the occupancy probability of each grid in the second occupancy grid map, the occupancy probability is fused, which has low fusion difficulty, low consumption and short time consumption. There is no need to perform feature fusion processing on data of different types and dimensions, and the method can solve the problem of difficulty in fusion due to differences in data types and dimensions when performing feature fusion on driving data from different sources.

[0083] In some embodiments, based on the above embodiments, the above step S302 "determine the second area of ​​interest of the vehicle at the second moment according to the first area of ​​interest and the driving data" may include: Figure 4 The following steps S3021 to S3023 are shown:

[0084] Step S3021: determining a movable range corresponding to the first focus area according to the first focus area.

[0085] Exemplarily, the first focus area of ​​the vehicle at the first moment is as follows: Figure 5 The solid line frame area shown in (a) above, the movable range corresponding to the first focus area is as follows: Figure 5 The first area of ​​interest is the surrounding area that the vehicle is concerned about at the first moment and is related to the vehicle's passage. The movable range corresponding to the first area of ​​interest is the area where the vehicle is allowed to move under the premise that the first area of ​​interest remains unchanged. When the vehicle moves outside the movable range, the area of ​​interest of the vehicle changes and the first area of ​​interest needs to be updated.

[0086] After the drivable area determination device determines the first area of ​​interest, the concentric area within the first area of ​​interest may be used as the movable range. Exemplarily, the size of the movable range may be slightly larger than the size of the sub-area corresponding to a sub-image. For example, the movable range may be as follows: Figure 5 The area within the dotted box shown in (a).

[0087] Step S3022: Determine the position of the vehicle at the second moment according to the position data in the driving data.

[0088] The driving data includes the position data detected by the position detector in the data acquisition device. Exemplarily, the position data may be the coordinates of the vehicle in the world coordinate system, specifically the coordinates of the center point of the rear axle of the vehicle in the world coordinate system, and the coordinates of the center point of the rear axle of the vehicle in the world coordinate system are used as the position of the vehicle at the second moment.

[0089] Step S3023: In response to the position of the vehicle at the second moment being within the movable range, determining a second area of ​​interest according to the first area of ​​interest.

[0090] After determining the position of the vehicle at the second moment, the device for determining the drivable area determines whether it is within the movable range corresponding to the first area of ​​interest. Figure 5 As shown in (b) in FIG, if the position of the vehicle at the second moment is within the movable range, it means that the distance moved by the vehicle between the first moment and the second moment is small and does not exceed the movable range corresponding to the first focus area. At this time, it is considered that the focus area of ​​the vehicle has not changed, so there is no need to update the focus area. The first focus area is determined as the second focus area. The obtained second focus area is as follows Figure 5 As shown in (c) in .

[0091] like Figure 5 As shown in (d) in the figure, if the position of the vehicle at the second moment is outside the movable range, it means that the vehicle has moved a large distance between the first moment and the second moment and exceeds the movable range corresponding to the first focus area. At this time, it is considered that the focus area of ​​the vehicle has changed and needs to be updated. Figure 4 The following steps S3024 to S3026 are shown to determine the second region of interest.

[0092] Step S3024, in response to the position of the vehicle at the second moment being not within the movable range, determining a target sub-region corresponding to the position of the vehicle at the second moment according to the sub-regions included in the first focus area and the position of the vehicle at the second moment.

[0093] For example, when the position of the vehicle at the second moment is not within the movable range, the target sub-region corresponding to the position of the vehicle at the second moment can be determined based on the sub-regions included in the first focus area and the position of the vehicle at the second moment. Figure 5 As shown in (d) in FIG. 1 , the position of the vehicle at the second moment is in the sub-region 501 , and the sub-region 501 is used as the target sub-region.

[0094] Step S3025: determine the center point of the second focus area according to the target sub-area.

[0095] For example, the center point of the target sub-region may be used as the center point of the second focus region. Figure 5 As shown in (d) in FIG. 5 , the center point P of the sub-region 501 is used as the center point of the second focus region.

[0096] Step S3026: Determine a second region of interest based on the center point and the first region of interest.

[0097] The device for determining the drivable area determines the second area of ​​interest at the second moment according to the center point P and the first area of ​​interest at the first moment. The obtained second area of ​​interest is as follows: Figure 5 As shown in (e) in .

[0098] In the disclosed embodiment, it is determined whether the vehicle's area of ​​interest has changed according to the vehicle's position in the driving data at the second moment. If it has not changed, the first area of ​​interest at the first moment is used as the second area of ​​interest at the second moment; if it has changed, the second area of ​​interest at the second moment is determined according to the center point of the sub-region where the vehicle position is located. When the vehicle moves a small distance from the first moment to the second moment, the first area of ​​interest is not updated. When the vehicle moves a large distance from the first moment to the second moment, the area of ​​interest is updated. There is no need to update the area of ​​interest in real time following the real-time position of the vehicle, which can reduce the update frequency of the area of ​​interest.

[0099] In some embodiments, based on the above embodiments, the above step S303 "determining the occupancy probability of each grid in the second occupancy grid map corresponding to the second area of ​​interest according to the driving data and the occupancy probability of each grid in the first occupancy grid map" may include Figure 6 The following steps S3031 and S3032 are shown:

[0100] Step S3031: determining the update probability of each grid in the second occupancy grid map corresponding to the second focus area according to the driving data.

[0101] Exemplarily, the device for determining the drivable area can determine the probability of each grid being occupied in the second occupied grid map based on the different source data included in the driving data at the second moment, and obtain the probability corresponding to each source data; then fuse the probabilities corresponding to each source data to determine the update probability of each grid corresponding to the driving data at the second moment.

[0102] The vehicle's driving data at the second moment (denoted as z) includes image data collected by the on-board camera (denoted as z1), millimeter-wave radar data collected by the millimeter-wave radar (denoted as z2), laser radar data collected by the laser radar (denoted as z3), and ultrasonic radar data collected by the ultrasonic radar (denoted as z4). Figure 7 As shown, the second grid map occupies the i-th grid in the m-th sub-map (denoted as m i ) as an example, the device for determining the drivable area determines the grid m according to the image data z1. i The first probability of being occupied is denoted as p1(m i |z1), determine the grid m according to the millimeter wave radar data z2 i The second probability of being occupied is denoted as p2(m i |z2), determine the grid m based on the lidar data z3 i The third probability of being occupied is denoted as p3(m i |z3), determine the grid m according to the ultrasonic radar data z4 i The fourth probability of being occupied is denoted as p4(m i |z4); Then, the first probability p1(m i |z1), the second probability p2(m i |z2), the third probability p3(m i |z3), and the fourth probability p4(m i |z4) is fused to obtain the grid m corresponding to the driving data z at the second moment i The update probability is denoted as l t (m i |z). Specifically, update probability l t (m i |z) can be fused using formula (1):

[0103]

[0104] Wherein, m is the sub-image number, ranging from 0 to 8. i is the grid number, ranging from 0 to N-1, and N is the number of grids in a single sub-image. t represents the second moment.

[0105] Alternatively, in other optional embodiments, the fusion weight of each sensor can be set according to the confidence of the perception results of different sensors. For example, the fusion weights of the vehicle camera, millimeter wave radar, laser radar and ultrasonic radar are a1, a2, a3 and a4 respectively. The probability l is updated by combining the fusion weights a1, a2, a3 and a4. t (m i |z) can be fused using formula (2):

[0106]

[0107] In the disclosed embodiment, when determining the update probability, the probability of occupation of each grid determined by different source data is fused. The fusion has low difficulty, low consumption, and short time. There is no need to perform feature fusion processing on data of different types and dimensions. It can solve the problem of difficulty in fusion due to differences in data types and dimensions when performing feature fusion on driving data from different sources.

[0108] Step S3032: determining the occupancy probability of each grid in the second occupancy grid map corresponding to the second focus area according to the update probability of each grid in the second occupancy grid map and the occupancy probability of each grid in the first occupancy grid map.

[0109] Exemplarily, after obtaining the update probability of each grid in the second occupation grid map and the occupation probability of each grid in the first occupation grid map, the device for determining the drivable area fuses the update probability and occupation probability of the grids corresponding to the same spatial area in the second occupation grid map and the first occupation grid map to obtain the occupation probability of each grid in the second occupation grid map corresponding to the second area of ​​interest.

[0110] When the second area of ​​interest is consistent with the first area of ​​interest, the combination order of the sub-image numbers in the second occupied grid map is consistent with the combination order of the sub-image numbers in the first occupied grid map, the spatial area corresponding to the ith grid in the mth sub-image in the second occupied grid map and the ith grid in the mth sub-image in the first occupied grid map is the same area, and the spatial area corresponding to the ith grid in the mth sub-image in the second occupied grid map is the same area. i The occupation probability l t (m i ) can be calculated according to formula (3):

[0111] l t (m i )=l t (m i |z)+l t-1 (m i ) (3);

[0112] Among them, t-1 represents the first moment. t-1 (m i ) represents the grid m in the first occupied grid map corresponding to the first moment i The probability of occupation.

[0113] When the second region of interest is inconsistent with the first region of interest, the combination order of the sub-image numbers in the second occupancy grid map is inconsistent with the combination order of the sub-image numbers in the first occupancy grid map. Since the storage space of the atomic images that do not need to be maintained in the first occupancy grid map is reset, for the reassembled sub-images, such as Figure 2 The sub-graphs 0 and 1 shown in (d) Figure 3 Kazuko Figure 6 , the occupancy probability l of each grid in the three sub-graphs t-1 (m i ) can be an initial value (such as 0.5), and the second occupies the grid m in the three sub-maps in the grid map. i The occupation probability l t (m i )=l t (m i |z)+l t-1 (m i )=l t (m i |z)+0.5. For sub-graphs whose storage space is not reset, that is, sub-graphs that are not reassembled, such as Figure 2 The sub- Figure 1 ,son Figure 2 ,son Figure 4 ,son Figure 5 ,son Figure 7 Kazuko Figure 8 , the occupancy probability l of each grid in the six sub-graphs t-1 (m i ) is the occupancy probability of each grid in the first occupied grid map calculated at the first moment and stored in the storage space corresponding to each sub-graph, and the occupancy probability of each grid in the six sub-graphs in the second occupied grid map is i The occupation probability l t (m i )=l t (m i |z)+l t-1 (m i ).

[0114] In the disclosed embodiment, the drivable area determination device determines the occupancy probability of each grid in the second occupied grid map by fusing the occupancy probability of the grid determined based on driving data at different times and / or from different sources. The fusion has low difficulty, low consumption, and short time consumption, and can quickly determine the occupancy probability of each grid in the second occupied grid map.

[0115] In some embodiments, based on the above embodiments, the above step S3031 "determining the update probability of each grid in the second occupancy grid map corresponding to the second focus area according to the driving data" may include: Figure 8 The following steps S30311 to S30313 are shown:

[0116] Step S30311, extract features from multiple source data in the driving data respectively to obtain feature contour point sets corresponding to each source data.

[0117] Exemplarily, the multi-source data in the driving data at the second moment may include image data collected by the vehicle-mounted camera, millimeter-wave radar data collected by the millimeter-wave radar, lidar data collected by the lidar, and ultrasonic radar data collected by the ultrasonic radar, etc.

[0118] When the device for determining the drivable area determines the probability of each grid being occupied in the second occupied grid map according to the data of different sources included in the driving data at the second moment, the device can first clean the data of each source in the driving data, and then extract features from the cleaned data to obtain a feature contour point set corresponding to each source data. The feature contour point set corresponding to each source data includes the feature contour points of the obstacles extracted from the source data.

[0119] In the disclosed embodiments, the dimensions and types of feature contour points extracted from different source data may be the same or different. For example, there is a pedestrian in front of the vehicle, and the feature contour point set extracted from the image data includes the feature contour points of the pedestrian image, wherein the feature contour points of the pedestrian image may be two-dimensional image features. The feature contour point set extracted from the millimeter-wave radar data includes the feature contour points of the electromagnetic wave signal reflected by the pedestrian, wherein the feature contour points of the electromagnetic wave signal reflected by the pedestrian may be three-dimensional electromagnetic wave features.

[0120] Data cleaning can improve the quality and availability of data, ensure the consistency and accuracy of data from different sources, and help improve the accuracy of the occupancy probability of each grid.

[0121] Step S30312, determining the grid coordinates of each feature contour point in each feature contour point set in the second occupied grid map corresponding to the second focus area.

[0122] After the device for determining the drivable area extracts the characteristic contour point set corresponding to each source data, the grid coordinates of each characteristic contour point in each characteristic contour point set in the second occupied grid map are determined. Exemplarily, the device for determining the drivable area can determine the grid coordinates through the following steps S3121 to S3123:

[0123] Step S3121, determining the world coordinates of each feature contour point in the feature contour point set in the world coordinate system.

[0124] The device for determining the drivable area can determine the world coordinates of each feature contour point in the feature contour point set corresponding to each source data based on the vehicle position detection information in the driving data, the position of each sensor in the vehicle in the data acquisition device, and the acquired parameters. Taking the feature contour point set corresponding to the image data as an example, the world coordinates of each feature contour point in the feature contour point set can be determined based on the world coordinates of the vehicle and the internal and external parameters of the vehicle-mounted camera.

[0125] Step S3122: Determine a mapping relationship between a grid map coordinate system of a second occupied grid map corresponding to the second focus area and a world coordinate system.

[0126] like Figure 2 As shown, the origin (upper left corner point) of the occupation grid map constructed by the device for determining the drivable area corresponds to the upper left corner point of the focus area. Based on this, the spatial area corresponding to each grid coordinate in the second occupation grid map in the second focus area is determined, and the mapping relationship between the grid map coordinate system and the world coordinate system is obtained.

[0127] Step S3123, according to the mapping relationship, the world coordinates of each feature contour point in the feature contour point set are converted into a grid map coordinate system to obtain the grid coordinates of each feature contour point in the feature contour point set.

[0128] According to the mapping relationship between the grid map coordinate system and the world coordinate system, the world coordinates of each feature contour point in the feature contour point set are converted to the grid map coordinate system to obtain the grid coordinates of each feature contour point. Fig. 9 is a schematic diagram of grid coordinates of each feature contour point in a grid map coordinate system provided by an exemplary embodiment of the present disclosure, wherein Fig. 9 (a) is a schematic diagram of the grid coordinates of each feature contour point in the feature contour point set corresponding to the image data. Fig. 9 (b) is a schematic diagram of the grid coordinates of each feature contour point in the feature contour point set corresponding to the millimeter wave radar data. Fig. 9 (c) in FIG. 1 is a schematic diagram of the grid coordinates of each feature contour point in the feature contour point set corresponding to the laser radar data. Fig. 9 (d) in FIG. 1 is a schematic diagram of the grid coordinates of each feature contour point in the feature contour point set corresponding to the ultrasonic radar data. Fig. 9 The middle grid 901 is a mapping grid corresponding to the world coordinates of the vehicle.

[0129] In the disclosed embodiment, the grid coordinates of each feature contour point in each feature contour point set in the second occupied grid map are determined through the mapping relationship between the world coordinate system and the grid map coordinate system, so that accurate grid coordinates can be obtained, which helps to obtain a more accurate drivable area.

[0130] Step S30313, determining the update probability of each grid in the second occupied grid map according to the grid coordinates of each feature contour point in each feature contour point set.

[0131] For each characteristic contour point, it is regarded as an observation of the grid, and the above formula (1) is used to calculate the probability that the grid corresponding to the observation is occupied to obtain the update probability. Exemplarily, the device for determining the drivable area can determine the update probability according to the following steps S3131 to S3133:

[0132] Step S3131 , in the first coordinate axis direction of the grid map coordinate system, interpolation is performed based on each feature contour point in each feature contour point set to obtain a plurality of feature interpolation contour points with continuous grid coordinates in the first coordinate axis direction.

[0133] The grids corresponding to the characteristic contour points determined according to step S30312 are discrete. In order to obtain a closed curve and improve the accuracy of the drivable area, the grids are interpolated in the disclosed embodiment to determine continuous characteristic interpolation contour points. When interpolating, the first coordinate axis direction of the grid map coordinate system can be interpolated first. Exemplarily, the first coordinate axis direction of the grid map coordinate system can be the MX direction. First, the characteristic contour points are interpolated in the MX direction to ensure that the grid coordinates in the MX direction are continuous and can form a closed line segment at both ends.

[0134] Fig.10 is a schematic diagram of the interpolation principle in the method for determining the drivable area provided by an exemplary embodiment of the present disclosure, and the mapping of the characteristic contour points occupying a 6*6 grid block in the grid map is as follows Fig.10 As shown in (a) in the figure, the grid coordinates of the corresponding feature contour points are (5,0), (4,1), (2,2) and (0,5). Among them, the feature contour points (5,0) and (4,1) are continuous in the MX direction and do not need to be interpolated. The feature contour points (4,1) and (2,2), (2,2) and (0,5) are not continuous in the MX direction. According to the feature contour points (4,1) and (2,2), interpolation is performed in the MX direction to obtain the interpolated contour point (3,1). According to the feature contour points (2,2) and (0,5), interpolation is performed in the MX direction to obtain the interpolated contour point (1,3). The interpolation result is shown in the figure below. Fig.10 As shown in (b), the original feature contour points and the inserted interpolation contour points are used as feature interpolation contour points to obtain multiple feature interpolation contour points (5,0), (4,1), (3,1), (2,2), (1,3) and (0,5) with continuous grid coordinates in the MX direction.

[0135] Step S3132, in the second coordinate axis direction of the grid map coordinate system, interpolation is performed according to a preset format and multiple characteristic interpolation contour points to obtain multiple characteristic target contour points whose grid coordinates are continuous in both the first coordinate axis direction and the second coordinate axis direction.

[0136] Exemplarily, the second coordinate axis direction of the grid map coordinate system may be the MY direction. After the interpolation in the MX direction is completed, the interpolation in the MY direction is continued to ensure that the grid coordinates in the MY direction are also continuous and can form a closed line segment. In this way, multiple characteristic target contour points that are continuous in both the MX direction and the MY direction are obtained.

[0137] See also Fig.10 In (b), among the characteristic interpolation contour points (5,0), (4,1), (3,1), (2,2), (1,3) and (0,5), (5,0), (4,1), (3,1), (2,2) and (1,3) are continuous in the MY direction and do not need interpolation. (1,3) and (0,5) are discontinuous in the MY direction. Interpolation is performed in the MY direction based on (1,3) and (0,5) to obtain the interpolation contour point (1,4). The interpolation result is shown in the figure below. Fig.10 As shown in (c) in the figure, the original feature interpolation contour points and the inserted interpolation contour points are used as feature target contour points, and multiple feature target contour points (5,0), (4,1), (3,1), (2,2), (1,3), (1,4) and (0,5) with continuous grid coordinates in the MX and MY directions are obtained.

[0138] According to the above interpolation method, Fig. 9 Interpolate (a) to (d) in , and obtain multiple characteristic target contour points whose grid coordinates are continuous in the first coordinate axis direction and the second coordinate axis direction. Fig. 9 The target contour points obtained by interpolation of (a) are as follows Fig.11 shown.

[0139] Step S3133: determining the update probability of each grid in the second occupied grid map according to each characteristic target contour point.

[0140] The characteristic target contour points include characteristic contour points and interpolation contour points. The update probability of the grid corresponding to the characteristic contour points can be directly fused according to the probabilities corresponding to each source data in the driving data, and the update probability of the grid corresponding to the interpolation contour points can be interpolated according to the fused probabilities.

[0141] In some embodiments, determining the update probability of each grid in the second occupancy grid map according to each characteristic target contour point may include the following steps:

[0142] 1) Determine the attribute information of each feature target contour point.

[0143] Exemplarily, the attribute information of the characteristic target contour point is used to determine the probability that the characteristic target contour point corresponds to a grid. The attribute information may include image feature attributes, millimeter wave feature attributes, laser feature attributes, ultrasonic feature attributes, and interpolation attributes.

[0144] 2) Determine the update probability of each grid in the second occupied grid map according to the attribute information of each characteristic target contour point and a preset probability table.

[0145] Among them, the preset probability table stores the prior probabilities of different types of data. Prior probability refers to the confidence level that the grid corresponding to the current data feature value is occupied based on past experience and analysis. The prior probability that the grid corresponding to the feature target contour points of different attributes is occupied can be obtained by looking up the table.

[0146] For feature target contour points whose attribute information is image feature attributes, millimeter wave feature attributes, laser feature attributes, and ultrasonic feature attributes, the preset probability table can be queried based on the attribute information of the feature target contour points to determine the prior probabilities corresponding to the source data of different attributes. Then, based on the feature values ​​of the feature contour points corresponding to each source data and the queried prior probabilities, the probability values ​​of the grids corresponding to the feature contour points being occupied are calculated; the probability values ​​of different source data corresponding to each grid are fused to obtain the update probabilities of each grid corresponding to the feature contour points.

[0147] To occupy grid m in the grid map i For example, according to the prior probability of image data z1 and image type data, the grid m is determined i The first probability value p1(m i |z1), determine the grid m according to the prior probability of the millimeter wave radar data z2 and the millimeter wave type data i The second probability value p2(m i |z2), determine the grid m based on the prior probability of the lidar data z3 and the laser type data i The third probability value p3(m i |z3), determine the grid m according to the prior probability of ultrasonic radar data z4 and ultrasonic type data i The fourth probability value p4(m i |z4); Then, the first probability value p1(m i |z1), the second probability value p2(m i |z2), the third probability value p3(m i |z3), and the fourth probability value p4(m i |z4) to fuse and obtain the grid m in the second occupied grid map i The update probability l t (m i |z). For example, Fig.10 The update probabilities of the feature contour points (5,0), (4,1), (2,2) and (0,5) shown in (c) can be calculated in this way. The calculated update probabilities of the four grids are respectively recorded as l t50 , l t41 , l t22 and l t05 .

[0148] For feature target contour points whose attribute information is interpolation attributes, the update probability of the interpolation contour points can be calculated based on the probability that the grid based on the interpolation is occupied. For example, Fig.10 The updated probability l of the interpolated contour points (3,1), (1,3) and (1,4) shown in (c) t31 , l t13 and l t14 According to l t41 , l t22 and l t05 Calculated. Among them, l t31 According to l t41 and l t22 Interpolation is obtained, l t13 According to l t22 and l t05 Interpolation is obtained, l t14 According to l t13 and l t05 The specific interpolation calculation method can be an average value method or a fitting value method, which is not limited in the embodiment of the present disclosure.

[0149] Through the above method, the update probability of each grid in the second occupied grid map is calculated.

[0150] In the embodiment of the present disclosure, by interpolating the feature contour points in the first coordinate axis and the second coordinate axis direction of the grid map coordinate system, continuous grids in the first coordinate axis and the second coordinate axis direction are obtained, and based on the continuous grids, the update probability of each grid in the second occupied grid map where the corresponding feature contour points exist and each interpolated grid is determined, thereby providing data support for calculating the occupancy probability of each grid in the second occupied grid map.

[0151] In some embodiments, when storing each feature contour point in the storage space corresponding to each sub-image, each feature contour point can be stored directly according to the grid coordinates, or steps S3131 and S3132 can be performed on each discrete feature contour point for interpolation. After the interpolation is completed, multiple feature target contour points whose grid coordinates are continuous in the first coordinate axis direction and the second coordinate axis direction are stored in a compressed manner.

[0152] Fig.12 is a schematic diagram of storing data in a compressed manner in a method for determining a drivable area provided by an exemplary embodiment of the present disclosure, such as Fig.12As shown in (a) in the figure, when compressed storage is used for storage, before compression, the MX and MY coordinate values ​​of the previous and next points during traversal are interpolated in the MX and MY axis directions, which are reflected in the Start_Length and End_Length data. The contour formed in this way is a completely closed figure that is continuous in both the MX and MY axes.

[0153] When storing the Yth row of the mth sub-graph, traverse from the first grid. When traversing to an occupied grid, fill it with X, indicating that there is an obstacle in the space area corresponding to the grid. Fill the grid between two occupied grids with O, indicating that the space area corresponding to the grid is passable. In each sub-graph, the area is filled according to the first and last points of each row. Compared with the ray-trace method of filling the area in the general grid graph, it can reduce the consumption of the processor when calculating the diagonal grid.

[0154] like Fig.12 As shown in (a), the first grid filled with X in row Y is marked as the starting position of the current row data, expressed as X_Start; the number of continuous grids filled with X in the first segment is marked as the starting segment length of the current row data, expressed as Start_Length; the last grid filled with X is marked as the ending position of the current row data, expressed as X_End; the number of continuous grids filled with X in the last segment is marked as the ending segment length of the current row data, expressed as End_Length. If no occupied grid is detected when traversing the current row data to the last grid, it means that there is no obstacle in the space area corresponding to the grid of the current row, and the grid of this row is filled with Free.

[0155] For example, in the process of updating the occupied grid map, std::vector can be used <freepointindex>The feature contour point information is compressed in the format of . The interpolated feature contour points are projected onto any sub-map occupying the grid map and encapsulated into std::vector <freepointindex>.

[0156]

[0157] For example, if there are multiple feature contour points in the same row of a sub-graph, such as Fig.12 As shown in (b), there are feature contour points 1, 2, 3 and 4 in the same row. At this time, according to the positional relationship of the feature contour points, the grid filling O between feature contour points 1 and 2 and the grid filling O between feature contour points 3 and 4 can be divided into an inner area 5 between feature contour points 1 and 2, a area 6 between feature contour points 3 and 4, and an outer area 7 between feature contour points 2 and 3 (not shown in the figure). The inner area is then abstracted into a FreePointIndex structure. Therefore, the same row of a sub-image (with the same Y value) may have multiple segments of compressed data.

[0158] In the disclosed embodiment, since the occupied grid is sparse, compression is preferred for storage to reduce the amount of data. In addition, when storing, each sub-graph corresponds to its own storage space, and searching for data according to the storage space corresponding to each sub-graph can improve the search speed compared to searching for data according to the entire occupied grid map.

[0159] In some embodiments, based on the above embodiments, the above step S304 "determine the drivable area in the second area of ​​interest according to the occupancy probability of each grid in the second occupancy grid map" may include: Fig.13 The following steps S3041 and S3042 are shown:

[0160] Step S3041: Determine the boundary grid of the drivable area according to the occupancy probability of each grid in the second occupancy grid map.

[0161] The boundary grid is a grid connecting the passable area and the obstacle area, and is usually located at the edge of the drivable area. Exemplarily, the drivable area determination device can traverse the grid map when determining the boundary grid: starting from the upper left corner, traverse each grid in the occupied grid map, and determine whether it belongs to an occupied grid based on the occupation probability of each grid; when the grid on one side adjacent to a grid is a non-occupied grid and the grid on the other side is an occupied grid, the grid is determined as a boundary grid.

[0162] Exemplarily, when traversing, the entire occupied grid map may be traversed row by row and column by column, or each sub-map in the occupied grid map may be traversed row by column. Other traversal methods may also be used, which are not limited in the embodiments of the present disclosure.

[0163] Step S3042: Determine a drivable area within the second area of ​​interest according to the boundary grid.

[0164] The device for determining the drivable area may connect boundary grids among all grids occupying the grid map to obtain the drivable area within the second area of ​​interest.

[0165] Exemplarily, since the grids are interpolated in the above embodiment, the determined boundary grids enclose a closed curve, and the area within the curve is the drivable area. The drivable area is composed of grids within the curve enclosed by the boundary grids, and the grids within the curve enclosed by the boundary grids can be marked as the drivable area.

[0166] In some embodiments, the device for determining the drivable area can be based on Fig.14 The following steps S30421 and S30424 are shown to determine the drivable area:

[0167] Step S30421, determining a mapping grid of the vehicle in a second occupation grid map according to the position data in the driving data.

[0168] For example, the device for determining the drivable area may determine the mapping grid of the vehicle in the second occupancy grid map according to the position data representing the vehicle coordinates in the driving data. Fig. 9 Grid 901 shown in .

[0169] Step S30422: Determine resampling information according to the yaw angle data in the driving data.

[0170] In order to speed up data search and extraction, a pre-stored sampling query table is adopted. According to the mapping grid and yaw angle (Yaw) corresponding to the vehicle position, the table information is traversed in sequence to obtain the contour point information. The sampling in the grid map has a Yaw change of 0-360 degrees. According to the vehicle Yaw, multiple sampling query tables can be pre-stored, for example, 180 grid query tables are stored (one table is stored every two degrees). When searching and extracting later, it is only necessary to query the corresponding pre-stored table according to the current Yaw angle of the vehicle to determine the resampling information, without the need for real-time calculation, so the extraction speed can be accelerated.

[0171] Step S30423, resample the boundary grid according to the mapping grid and the resampling information to obtain the contour grid.

[0172] For example, Fig.15 In the scenario shown, the resampling information determined based on the yaw angle in the driving data can be resampled in a first manner for 180° in front and behind the vehicle and in a second manner for 50° on both sides of the vehicle. The occupancy grid map is resampled according to the first manner and the second manner to obtain a contour grid.

[0173] Step S30424: Determine the drivable area within the second area of ​​interest based on the contour grid.

[0174] The drivable area in the second area of ​​interest is generated based on the resampled contour grid. Determining the drivable area based on the resampled contour grid in the grid map can reduce the amount of data of the determined drivable area, and can more quickly transmit the data of the drivable area to the subsequent processing module to more quickly implement path planning.

[0175] Since the grid map data volume is large, directly transmitting the map data during data processing occupies too much bandwidth. In order to improve the transmission speed, in the embodiment of the present disclosure, after constructing the grid map, the contour grid of the drivable area can be resampled on the grid map. The drivable area determined based on the resampled contour grid can greatly reduce the amount of data, which is beneficial to reducing the subsequent processing calculation amount and improving the calculation speed.

[0176] Exemplary Devices

[0177] The method for determining a drivable area provided in the above embodiment may be executed by a drivable area determination device, and specifically may be executed by a drivable area determination means in the drivable area determination device. Fig.16 FIG. 1 is a schematic diagram of the structure of a device for determining a drivable area provided by an exemplary embodiment of the present disclosure. Fig.16 As shown, the drivable area determination device 1600 may include:

[0178] A first determination module 1601 is used to determine a first focus area of ​​the vehicle at a first moment, an occupancy probability of each grid in a first occupancy grid map corresponding to the first focus area, and driving data of the vehicle at a second moment; wherein the second moment is later than the first moment;

[0179] A second determining module 1602, configured to determine a second area of ​​interest of the vehicle at the second moment according to the first area of ​​interest and the driving data;

[0180] A third determination module 1603 is used to determine the occupancy probability of each grid in the second occupancy grid map corresponding to the second focus area according to the driving data and the occupancy probability of each grid in the first occupancy grid map;

[0181] The fourth determination module 1604 is configured to determine a drivable area within the second region of interest according to the occupancy probability of each grid in the second occupancy grid map.

[0182] See also Fig.17 In some embodiments, the second determining module 1602 may include:

[0183] A first determining unit 16021 is configured to determine a movable range corresponding to the first focus area according to the first focus area;

[0184] A second determining unit 16022, configured to determine a position of the vehicle at the second moment according to the position data in the driving data;

[0185] The third determining unit 16023 is configured to determine the second area of ​​interest according to the first area of ​​interest in response to the position of the vehicle at the second moment being within the movable range.

[0186] See also Fig.17 In some embodiments, the second determining module 1602 may further include:

[0187] a fourth determining unit 16024, configured to determine, in response to the position of the vehicle at the second moment being not within the movable range, a target sub-region corresponding to the position of the vehicle at the second moment according to the sub-regions included in the first focus area and the position of the vehicle at the second moment;

[0188] A fifth determining unit 16025 is configured to determine a center point of the second focus area according to the target sub-area;

[0189] The sixth determining unit 16026 is configured to determine the second region of interest based on the center point and the first region of interest.

[0190] See also Fig.17 In some embodiments, the third determining module 1603 may include:

[0191] A seventh determining unit 16031 is used to determine, according to the driving data, an update probability of each grid in a second occupancy grid map corresponding to the second focus area;

[0192] The eighth determining unit 16032 is configured to determine the occupancy probability of each grid in the second occupancy grid map corresponding to the second focus area according to the update probability of each grid in the second occupancy grid map and the occupancy probability of each grid in the first occupancy grid map.

[0193] In some embodiments, the seventh determining unit 16031 may also be used to:

[0194] Extracting features from multiple source data in the driving data respectively to obtain feature contour point sets corresponding to each source data;

[0195] Determine the grid coordinates of each characteristic contour point in each characteristic contour point set in a second occupied grid map corresponding to the second focus area;

[0196] The update probability of each grid in the second occupied grid map is determined according to the grid coordinates of each characteristic contour point in each characteristic contour point set.

[0197] In some embodiments, the seventh determining unit 16031 may also be used to:

[0198] Determine the world coordinates of each feature contour point in the feature contour point set in the world coordinate system;

[0199] Determine a mapping relationship between a grid map coordinate system of a second occupied grid map corresponding to the second area of ​​interest and the world coordinate system;

[0200] According to the mapping relationship, the world coordinates of each feature contour point in the feature contour point set are converted into the grid map coordinate system to obtain the grid coordinates of each feature contour point in the feature contour point set.

[0201] In some embodiments, the seventh determining unit 16031 may also be used to:

[0202] In the first coordinate axis direction of the grid map coordinate system, interpolation is performed according to each feature contour point in each feature contour point set to obtain a plurality of feature interpolation contour points with continuous grid coordinates in the first coordinate axis direction;

[0203] In the second coordinate axis direction of the grid map coordinate system, interpolation is performed according to a preset format and the plurality of characteristic interpolation contour points to obtain a plurality of characteristic target contour points whose grid coordinates are continuous in both the first coordinate axis direction and the second coordinate axis direction;

[0204] According to each of the characteristic target contour points, an update probability of each grid in the second occupancy grid map is determined.

[0205] In some embodiments, the seventh determining unit 16031 may also be used to:

[0206] Determining the attribute information of each of the characteristic target contour points;

[0207] The update probability of each grid in the second occupancy grid map is determined according to the attribute information of each characteristic target contour point and a preset probability table.

[0208] See also Fig.17 In some embodiments, the fourth determining module 1604 may include:

[0209] A ninth determining unit 16041, configured to determine a boundary grid of the drivable area according to the occupancy probability of each grid in the second occupancy grid map;

[0210] The tenth determining unit 16042 is configured to determine a drivable area within the second area of ​​interest according to the boundary grid.

[0211] In some embodiments, the tenth determining unit 16042 may also be used to:

[0212] determining a mapping grid of the vehicle in the second occupancy grid map according to the position data in the driving data;

[0213] Determining resampling information according to the yaw angle data in the driving data;

[0214] Resampling the boundary grid according to the mapping grid and the resampling information to obtain a contour grid;

[0215] A drivable area within the second area of ​​interest is determined according to the contour grid.

[0216] The beneficial technical effects corresponding to the exemplary embodiment of the present device can be found in the corresponding beneficial technical effects of the above exemplary method section, which will not be repeated here.

[0217] Exemplary Electronic Devices

[0218] Fig.18 is a schematic diagram of the composition structure of an electronic device provided by an exemplary embodiment of the present disclosure, such as Fig.18 As shown, the electronic device 1800 may include at least one processor 1801 and a memory 1802 .

[0219] The processor 1801 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 1800 to perform desired functions.

[0220] The memory 1802 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on a computer-readable storage medium, and the processor 1801 may execute one or more computer program instructions to implement the method for determining the drivable area and / or other desired functions of the various embodiments of the present disclosure described above.

[0221] In one example, the electronic device 1800 may further include: an input device 1803 and an output device 1804 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0222] The input device 1803 may also include, for example, a keyboard, a mouse, etc.

[0223] The output device 1804 can output various information to the outside, and may include, for example, a display, a speaker, a printer, a communication network and its connected remote output device, etc.

[0224] Of course, to simplify, Fig.18 Only some of the components related to the present disclosure in the electronic device 1800 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 1800 may also include any other appropriate components.

[0225] Exemplary computer program products and computer-readable storage media

[0226] In addition to the above-mentioned methods and devices, embodiments of the present disclosure may also provide a computer program product, including computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the method for determining the drivable area of ​​various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section.

[0227] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0228] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the processor executes the steps in the method for determining the drivable area of ​​various embodiments of the present disclosure described in the above “Exemplary Method” section.

[0229] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium is, for example, but not limited to, a system, device or device including electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0230] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and cannot be considered as necessary for each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, rather than limitation, and the above details do not limit the present disclosure to being implemented by adopting the above specific details.

[0231] Those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present disclosure claims and their equivalents, the present disclosure is also intended to include these modifications and variations.< / freepointindex> < / freepointindex>

Claims

1. A method for determining a drivable area, comprising: Determine a first focus area of ​​the vehicle at a first moment, an occupancy probability of each grid in a first occupancy grid map corresponding to the first focus area, and driving data of the vehicle at a second moment; wherein the second moment is later than the first moment; determining a second area of ​​interest for the vehicle at the second moment according to the first area of ​​interest and the driving data; Determining, according to the driving data and the occupancy probability of each grid in the first occupancy grid map, the occupancy probability of each grid in the second occupancy grid map corresponding to the second area of ​​interest; A drivable area within the second area of ​​interest is determined according to the occupancy probability of each grid in the second occupancy grid map.

2. The method according to claim 1, wherein: The determining, according to the first area of ​​interest and the driving data, a second area of ​​interest of the vehicle at the second moment, comprises: Determining a movable range corresponding to the first area of ​​interest according to the first area of ​​interest; Determining the position of the vehicle at the second moment according to the position data in the driving data; In response to the position of the vehicle at the second moment being within the movable range, the second area of ​​interest is determined according to the first area of ​​interest.

3. The method according to claim 2, wherein: The determining, according to the first area of ​​interest and the driving data, a second area of ​​interest of the vehicle at the second moment, further comprises: In response to the position of the vehicle at the second moment not being within the movable range, determining a target sub-region corresponding to the position of the vehicle at the second moment according to the sub-regions included in the first area of ​​interest and the position of the vehicle at the second moment; Determining a center point of the second focus area according to the target sub-area; The second region of interest is determined according to the central point and the first region of interest.

4. The method according to claim 1, wherein: The determining, based on the driving data and the occupancy probability of each grid in the first occupancy grid map, the occupancy probability of each grid in the second occupancy grid map corresponding to the second focus area comprises: Determining, based on the driving data, an update probability of each grid in a second occupancy grid map corresponding to the second area of ​​interest; The occupancy probability of each grid in the second occupancy grid map corresponding to the second region of interest is determined according to the update probability of each grid in the second occupancy grid map and the occupancy probability of each grid in the first occupancy grid map.

5. The method according to claim 4, wherein: The step of determining, based on the driving data, an update probability of each grid in a second occupancy grid map corresponding to the second focus area comprises: Extracting features from multiple source data in the driving data respectively to obtain feature contour point sets corresponding to each source data; Determine the grid coordinates of each characteristic contour point in each characteristic contour point set in a second occupied grid map corresponding to the second focus area; The update probability of each grid in the second occupied grid map is determined according to the grid coordinates of each characteristic contour point in each characteristic contour point set.

6. The method according to claim 5, wherein: The step of determining the grid coordinates of each characteristic contour point in each characteristic contour point set in the second occupied grid map corresponding to the second focus area comprises: Determine the world coordinates of each feature contour point in the feature contour point set in the world coordinate system; Determine a mapping relationship between a grid map coordinate system of a second occupied grid map corresponding to the second area of ​​interest and the world coordinate system; According to the mapping relationship, the world coordinates of each feature contour point in the feature contour point set are converted into the grid map coordinate system to obtain the grid coordinates of each feature contour point in the feature contour point set.

7. The method according to claim 6, wherein: The step of determining the update probability of each grid in the second occupied grid map according to the grid coordinates of each characteristic contour point in each characteristic contour point set comprises: In the first coordinate axis direction of the grid map coordinate system, interpolation is performed according to each feature contour point in each feature contour point set to obtain a plurality of feature interpolation contour points with continuous grid coordinates in the first coordinate axis direction; In the second coordinate axis direction of the grid map coordinate system, interpolation is performed according to a preset format and the plurality of characteristic interpolation contour points to obtain a plurality of characteristic target contour points whose grid coordinates are continuous in both the first coordinate axis direction and the second coordinate axis direction; According to each of the characteristic target contour points, an update probability of each grid in the second occupancy grid map is determined.

8. The method according to claim 7, wherein: Determining the update probability of each grid in the second occupied grid map according to each of the characteristic target contour points includes: Determining the attribute information of each of the characteristic target contour points; The update probability of each grid in the second occupancy grid map is determined according to the attribute information of each characteristic target contour point and a preset probability table.

9. The method according to any one of claims 1 to 8, wherein: The determining, according to the occupancy probability of each grid in the second occupancy grid map, a drivable area within the second area of ​​interest includes: determining a boundary grid of the drivable area according to the occupancy probability of each grid in the second occupancy grid map; A drivable area within the second area of ​​interest is determined according to the boundary grid.

10. The method according to claim 9, wherein: The step of determining the drivable area within the second area of ​​interest according to the boundary grid includes: determining a mapping grid of the vehicle in the second occupancy grid map according to the position data in the driving data; Determining resampling information according to the yaw angle data in the driving data; Resampling the boundary grid according to the mapping grid and the resampling information to obtain a contour grid; A drivable area within the second area of ​​interest is determined according to the contour grid.

11. A device for determining a drivable area, comprising: A first determination module is used to determine a first focus area of ​​the vehicle at a first moment, an occupancy probability of each grid in a first occupancy grid map corresponding to the first focus area, and driving data of the vehicle at a second moment; wherein the second moment is later than the first moment; A second determining module, configured to determine a second area of ​​interest of the vehicle at the second moment according to the first area of ​​interest and the driving data; a third determination module, configured to determine the occupancy probability of each grid in the second occupancy grid map corresponding to the second area of ​​interest based on the driving data and the occupancy probability of each grid in the first occupancy grid map; The fourth determination module is used to determine the drivable area in the second area of ​​interest according to the occupancy probability of each grid in the second occupancy grid map.

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

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

Images