Raster Map Generation Method and Device, Mobile Tool, Storage Medium

Abstract

The present invention discloses a method and device for generating a grid map. The method includes obtaining multi-line laser point cloud, and generating a two-dimensional scan data set according to the multi-line laser point cloud. Wherein, the two-dimensional scan data set is a set of two-dimensional data points generated based on scan points meeting a first preset condition, and the two-dimensional data points have height attributes; generating a refresh point set according to the two-dimensional scan data set, wherein the refresh point set is a set formed by two-dimensional data points meeting a second preset condition, and the two-dimensional data points in the refresh point set also have passable attributes; updating the grid map according to the refresh point set. The solution provided by the embodiments of the present invention can quickly and low-costly construct a grid map that can cover low objects and blind areas in a scenario of three-dimensional scanning based on a multi-line laser sensor, expanding the application scenario of the two-dimensional grid map and making it applicable to vehicle and robot operation scenarios with a certain height.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving, and particularly to a method for generating a grid map, a grid map generation device, a mobile tool, and a storage medium. Background Art

[0002] Autonomous driving technology is one of the currently widely concerned technical directions. Typical autonomous driving technologies mainly cover three main directions at present: perception technology, positioning and mapping technology, and decision-making and planning technology. In order to achieve autonomous movement, real-time positioning and map information are essential information sources for autonomous driving devices. Most robots and autonomous driving vehicles need to scan the surrounding environment and construct three-dimensional or two-dimensional maps for subsequent decision-making and planning, etc. Common types of maps include high-precision maps (HD maps), point cloud maps (Point cloud Maps), and grid maps (Grid maps).

[0003] Currently, the mainstream map construction methods are mainly of two types: one is to scan the environment through a single-line lidar and then perform simultaneous localization and mapping technology (SLAM) on the scan results to achieve online construction of a grid map; this approach is effective in scenarios with a small size and a small environmental area. For example, in a household scenario within 200 square meters, most robots adopt this map construction method; the other is to use the point cloud map obtained by multi-line laser scanning as the basic map data source and build a high-precision map based on this; this technology is also called SLAM technology, but its implementation details are more complex than those of two-dimensional grid maps. Since the map production and annotation efficiency of the high-precision map construction scheme based on multi-line lasers is much lower than that of the grid map construction scheme based on single-line lidar, the high-precision map construction scheme based on multi-line lasers is usually applied to large autonomous driving vehicles for carrying people and goods with high requirements.

[0004] For indoor low-speed autonomous driving vehicles or small commercial vehicles, since their size is significantly larger than that of a sweeping robot but not as complex as that of a manned vehicle, if the single-line laser grid map scheme or the multi-line laser high-precision map scheme is directly used on this type of autonomous driving vehicle, certain problems will occur, such as:

[0005] 1) For a floor cleaning robot, its shape is usually flat and circular, and the laser sensor is installed at the center or near the front of its upper surface. This way, the passable area detected by the laser is basically exactly the same as the passable area of the floor cleaning robot itself. However, in vehicles or robots with a certain height, if a single-line laser solution is used, the road surface or other height information cannot be detected, and such map information is not rich enough for navigation and is prone to causing safety problems.

[0006] 2) When directly using a multi-line laser sensor for environmental scanning in a low-speed autonomous vehicle, if the solution of making a high-precision map based on the point cloud map of a passenger vehicle is adopted, it is too complex for indoor vehicles or small commercial vehicles, and the cost is high and the burden is heavy.

[0007] Based on this, there is an urgent need to propose a more cost-effective map construction solution suitable for low-speed autonomous vehicles or small commercial vehicles indoors. Summary of the Invention

[0008] Embodiments of the present invention provide a grid map construction solution to solve the problem that directly using a single-line laser grid map solution in the prior art will cause a serious reduction in the map accuracy of vehicles or robots with a certain height, while directly using a multi-line laser high-precision map solution will cause an increase in map construction costs and computing power requirements, and thus cannot meet the map construction requirements of vehicles or robots in specific scenarios.

[0009] In a first aspect, embodiments of the present invention provide a grid map generation method, which includes:

[0010] Obtain multi-line laser point clouds, and generate a two-dimensional scan data set according to the multi-line laser point clouds, where the two-dimensional scan data set is a set of two-dimensional data points generated based on scan points that meet a first preset condition, and the two-dimensional data points have height attributes;

[0011] Generate a refresh point set according to the two-dimensional scan data set, where the refresh point set is a set formed by two-dimensional data points that meet a second preset condition, and the two-dimensional data points in the refresh point set also have passable attributes;

[0012] Update the grid map according to the refresh point set.

[0013] In a second aspect, embodiments of the present invention provide a grid map generation device, which includes:

[0014] A point cloud acquisition module for acquiring multi-line laser point clouds;

[0015] The first point set determination module is configured to generate a two-dimensional scan data set based on the multi-line laser point cloud, wherein the two-dimensional scan data set is a set of two-dimensional data points generated based on scan points meeting a first preset condition, and the two-dimensional data points have height attributes;

[0016] The second point set determination module is configured to generate a refresh point set based on the two-dimensional scan data set, wherein the refresh point set is a set formed by two-dimensional data points meeting a second preset condition, and the two-dimensional data points in the refresh point set also have passable attributes;

[0017] The map update module is configured to update the grid map based on the refresh point set.

[0018] In a third aspect, an embodiment of the present invention provides another grid map generation device, which includes:

[0019] A memory for storing executable instructions; and

[0020] A processor for executing the executable instructions stored in the memory, and the executable instructions, when executed by the processor, implement the method steps provided in the first aspect of the present invention.

[0021] In a fourth aspect, an embodiment of the present invention provides a mobile tool, which includes the grid map generation device according to the third aspect of the present invention.

[0022] In a fifth aspect, an embodiment of the present invention provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the steps of the method provided in the first aspect above.

[0023] In a sixth aspect, an embodiment of the present invention provides a computer program product, the computer program product includes a computer program stored on a non-volatile computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, the computer is made to execute the method provided in the first aspect above.

[0024] The beneficial effects of the embodiments of the present invention are as follows: The method provided by the embodiments of the present invention first converts the multi-line laser three-dimensional point cloud collected by the multi-line laser sensor into a two-dimensional scan data set, and then uses the generated two-dimensional scan data set to update the grid map, realizing the optimization of the traditional two-dimensional grid map update scheme, so that both three-dimensional scanning can be achieved through the multi-line laser sensor, and a grid map that can cover low objects and blind areas can be constructed quickly and at low cost, expanding the application scenario of the two-dimensional grid map, making it applicable to the operation scenarios of vehicles and robots with a certain height. Moreover, the solution of the present invention obtains the surrounding environment information based on multi-line laser scanning, so that the constructed two-dimensional grid map can not only avoid the visual blind area and model low objects, but also retain the moving object removal characteristic of the grid map. And since the three-dimensional grid is not directly constructed using the three-dimensional point cloud, the complexity and computational amount of the method of the embodiments of the present invention are significantly reduced compared with the three-dimensional map construction scheme based on multi-line laser scanning. Therefore, it can run in real time in the vehicle industrial computer environment without occupying too much computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0026] Figure 1 It is a schematic flow chart of a method for generating a grid map according to an embodiment of the present invention;

[0027] Figure 2 It is a schematic flow chart of a method for generating a two-dimensional scan data set according to multi-line laser point cloud according to an embodiment of the present invention;

[0028] Figure 3 It schematically shows the relationship between the height of the vehicle itself and the actual scanning situation of the multi-line laser sensor thereon and the display effect of the geometric model constructed based on this;

[0029] Figure 4 It schematically shows the conversion effect of generating a two-dimensional scan data set according to multi-line laser point cloud;

[0030] Figure 5 It schematically shows a schematic flow chart of a method for generating a refresh point set according to an embodiment of the present invention;

[0031] Figure 6 It schematically shows a schematic flow chart of a method for updating the grid map corresponding to the target area using the refresh point set according to an embodiment of the present invention;

[0032] Figure 7 Schematically shows the Figure 6 flow schematic diagram of the implementation method of step S122 in

[0033] Figure 8 Schematically shows the Figure 7 flow schematic diagram of the implementation method of step S1222 in

[0034] Figure 9 is a principle block diagram of a grid map generation device according to an embodiment of the present invention;

[0035] Figure 10 is a principle block diagram of a grid map generation device according to another embodiment of the present invention;

[0036] Figure 11 is a principle block diagram of a mobile tool according to an embodiment of the present invention;

[0037] Figure 12 is a structural schematic diagram of an embodiment of a grid map generation device of the present invention. Specific embodiments

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0039] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0040] The present invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media including storage devices.

[0041] In the present invention, "module", "device", "system", etc. refer to relevant entities applied to a computer, such as hardware, a combination of hardware and software, software, or software in execution. Specifically, for example, an element can be, but is not limited to, a process running on a processor, a processor, an object, an executable element, an execution thread, a program, and / or a computer. Also, an application program or a script program running on a server, and the server can both be elements. One or more elements can be in an execution process and / or thread, and the elements can be localized on one computer and / or distributed between two or more computers, and can be run by various computer-readable media. The elements can also communicate through local and / or remote processes according to a signal having one or more data packets, for example, a signal from data that interacts with another element in a local system, a distributed system, and / or interacts with other systems through a signal on a network in the Internet.

[0042] Finally, it should also be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprising" and "including" not only include those elements, but also include other elements not explicitly listed, or also include elements inherent to such a process, method, article, or device. Without more limitations, the elements defined by the statement "comprising..." do not exclude the existence of additional identical elements in the process, method, article, or device including the said elements.

[0043] The grid map generation method in the embodiments of the present invention can be applied to any product that needs to use positioning and mapping technologies, especially in autonomous driving vehicles and robots based on multi-line laser sensors with a certain height, so that these products can utilize the grid map generation scheme in the embodiments of the present invention to construct an optimized two-dimensional grid map to achieve full coverage of blind areas and low objects, and avoid marking obstacles in low objects and / or blind areas as passable areas. These products that need to use positioning and mapping technologies include, for example, but are not limited to mobile tools capable of achieving automatic driving or semi-automatic driving, such as autonomous driving vehicles (passenger cars, buses, large buses, trucks, lorries, etc.), unmanned sweeping machines, unmanned cleaning vehicles, robots, unmanned floor sweepers, floor-sweeping robots, and other robots. The present invention does not make any limitations in this regard.

[0044] Figure 1Schematically shows a grid map generation method according to an embodiment of the present invention. The execution subject of this method can be a positioning device or a controller on an autonomous vehicle, or a processor of a grid map generation device such as a smart tablet, a personal PC, a computer, a cloud server, etc., or a processor of an intelligent mobile platform such as an unmanned cleaning vehicle, an unmanned sweeping vehicle, a sweeping robot, an autonomous vehicle, a semi-autonomous device, etc. The embodiments of the present invention do not limit this. As Figure 1 shown, the method of the embodiment of the present invention includes:

[0045] Step S10: Obtain multi-line laser point cloud, and generate a two-dimensional scan data set according to the multi-line laser point cloud, wherein the two-dimensional scan data set is a set of two-dimensional data points formed based on scan points that meet a first preset condition, and the two-dimensional data points have height attributes;

[0046] Step S11: Generate a refresh point set according to the two-dimensional scan data set, wherein the refresh point set is a set formed by two-dimensional data points that meet a second preset condition, and the two-dimensional data points in the refresh point set are also assigned passable attributes;

[0047] Step S12: Update the grid map according to the refresh point set.

[0048] Among them, the multi-line laser point cloud in the embodiment of the present invention is information data in the surrounding environment obtained by a multi-line laser sensor such as a multi-line lidar, which at least includes a series of three-dimensional coordinates, and of course can further include some other information, such as auxiliary information such as reflectivity or color. The embodiment of the present invention refers to these surrounding environment information data obtained by the multi-line laser sensor as multi-line laser point cloud. It can be understood that these multi-line laser point clouds belong to three-dimensional point cloud data. In step S10, the embodiment of the present invention can obtain the multi-line laser point cloud through a multi-line laser sensor or other front-end modules. Since in mobile tools such as vehicles or robots with a certain height, the end points scanned by the multi-line laser sensor are not completely consistent with the passable area of the mobile tool itself, the two-dimensional grid algorithm based on single-line laser cannot be directly applied to the multi-line laser sensor. Therefore, in step S10, the embodiment of the present invention will first convert the obtained multi-line laser point cloud into a two-dimensional scan data set, so that the solution of the embodiment of the present invention can update the grid map based on the converted two-dimensional scan data set, so as to reduce the computational complexity of directly using three-dimensional point cloud data to construct a three-dimensional grid and generate a three-dimensional map on the basis of ensuring a higher coverage rate of environmental information by multi-line scanning.

[0049] As a preferred embodiment, in step S10, converting the multi-line laser point cloud into a two-dimensional scan data set can be achieved by means of a geometric model constructed based on the height of the mobile tool itself. Due to the limitation of the horizontal angle of laser sensors such as lidar, it is impossible to detect areas that are very close to itself and have a height lower than itself. Therefore, for vehicles or robots with a certain height, the laser sensors installed on them have scanning blind spots. Therefore, considering the height of the mobile tool itself, the laser end points scanned by the multi-line laser sensor do not exactly coincide with the passable area of the mobile tool itself. That is, the laser end points do not all fall on the ground, but there are blind spots and height differences. Based on this, the embodiments of the present invention preferably consider constructing a geometric model based on the height of the mobile tool to use this geometric model to determine the first ground contact point cloud of the multi-line laser point cloud in each azimuth angle direction, and generate the determined first ground contact point cloud into a two-dimensional scan data set. Thus, in step S10, the scan points that meet the first preset condition can refer to the first ground contact point cloud of the multi-line laser point cloud in each azimuth angle direction. Taking this as an example, Figure 2 Schematically shows a method for generating a two-dimensional scan data set according to a multi-line laser point cloud, as Figure 2 shown, the method includes the following steps:

[0050] Step S101: Sort the multi-line laser point cloud according to the azimuth angle, and sort the multi-line laser point cloud in each azimuth angle direction according to the distance;

[0051] Step S102: Determine the first ground contact point cloud of the multi-line laser point cloud in each azimuth angle direction;

[0052] Step S103: Generate a two-dimensional scan data set according to the determined first ground contact point cloud in each azimuth angle direction.

[0053] In order to clearly show the influence of the height of the mobile tool itself on the actual scanning result of the multi-line laser sensor, Figure 3 Taking the mobile tool as a vehicle as an example, schematically shows the relationship between the vehicle geometric model and the scanning situation of the multi-line laser sensor on it, as Figure 3As shown, since the vehicle 1 has a certain height H itself, and the laser sensor 2 itself has a horizontal angle limitation, in actual scanning, in the area M close to the vehicle 1, it will become the scanning blind area of the laser sensor 2. And the position where the laser end point scanned by the laser sensor 2 is located is not necessarily on the ground F, but there are two situations: on the ground F or on the obstacle O, as shown by the laser end points A1 and A2 in the figure. Therefore, considering the height of the vehicle itself, the passable area scanned by the multi-line laser sensor is actually a trapezoidal model 3. The bottom side of the trapezoidal model 3 is limited by the height H of the vehicle, the top side is limited by the ground clearance h of the point cloud at the first ground contact point on a certain scan line, one waist side is limited by the horizontal ground distance r from the vehicle to the first ground contact point, and the other waist side is limited by the scan distance s from the starting point of the scan line that defines the top side to the first ground contact point. Based on this, considering the influence of the actual height of the mobile tool on the scanning result, in the preferred embodiment of the present invention, when converting the multi-line laser point cloud into a two-dimensional scan data set, the actual position of the laser end point of each scan line will be considered, so as to perform two-dimensional conversion based on the actual position of the laser end point of the scan line of the multi-line laser to determine the first ground contact point cloud on each scan line (that is, the point cloud where the laser end point falls on the obstacle, such as Figure 3 the end point A1 in

[0054] is the first ground contact point cloud of this scan line).

[0055] Since the laser beam bundles scanned by the multi-line laser sensor, that is, the scan lines, include multiple lines in different directions, in order to determine the first ground contact point cloud in each direction, the embodiment of the present invention will first sort all the multi-line laser point clouds according to the azimuth angle through step S101 to determine all the scan directions and ensure full coverage in all directions. And after the azimuth angle sorting, the embodiment of the present invention will also perform distance sorting on the multi-line laser point clouds at each azimuth angle to quickly determine the first ground contact point cloud in each direction. Among them, in step S101, the distance sorting of the multi-line laser point cloud is specifically to sort the laser point clouds in each azimuth angle direction according to the horizontal distance from each point cloud on the multi-line laser point cloud to the vehicle, so that the laser point clouds at each azimuth angle are arranged in the order from near to far. In this way, by traversing the laser point clouds in each azimuth angle direction, the first ground contact point cloud in each direction can be quickly determined.Among them, in step S102, the detection of the first ground point cloud is performed for each sorted azimuth angle, so as to be able to determine the first ground point cloud in the direction of each azimuth angle and avoid omission. As a preferred embodiment, step S102 can be specifically implemented by determining the ground parameters, that is, calculating the first ground point cloud in the direction of each azimuth angle based on the determined ground parameters. Among them, determining the ground parameters can be specifically implemented as presetting a fixed ground height according to empirical values. Thus, in step S102, the first ground point cloud can be determined according to the ground heights of the laser point clouds in the direction of each sorted azimuth angle. Exemplarily, the ground heights of the laser point clouds in the direction of each azimuth angle can be sequentially compared with the preset ground height. Taking the example that in step S101, the laser point clouds are sorted by distance from near to far, when the first point cloud with a ground height greater than the preset ground height is found, this point cloud is determined as the first ground point cloud in this direction. In other embodiments, determining the first ground point cloud in each direction can also be achieved by first extracting the ground point cloud from the multi-line laser point cloud by methods such as point cloud segmentation. Specifically, the laser point clouds in each direction can be respectively compared with the ground point cloud to determine the first ground point cloud in this direction according to the comparison result. Among them, the implementation methods such as point cloud segmentation for extracting the ground point cloud from the multi-line laser point cloud can be implemented with reference to the prior art, and the embodiments of the present invention will not elaborate on this. Based on the ground point cloud, the first ground point cloud in each direction can be determined. Exemplarily, the height mean of the extracted ground point cloud can be determined by means such as mean calculation, and the ground heights of the laser point clouds in each direction are compared with this height mean to find the first ground point cloud in each direction.

[0056] In step S103, generating a two-dimensional point cloud dataset based on the first ground contact points in each determined direction can be specifically achieved by generating the first ground contact point clouds in each azimuth angle direction into corresponding two-dimensional data points respectively, and forming the two-dimensional data points corresponding to the first ground contact point clouds in each azimuth angle direction into a two-dimensional scan dataset. Among them, generating the two-dimensional data points corresponding to each first ground contact point cloud can be specifically to sequentially obtain the ground clearance height, azimuth angle, and distance of the first ground contact point cloud in each azimuth angle direction, and jointly define the two-dimensional data points corresponding to each first ground contact point cloud according to the ground clearance height, azimuth angle, and distance of each first ground contact point cloud. Exemplarily, the two-dimensional data points corresponding to the first ground contact point cloud with a ground clearance height of h, an azimuth angle of θ, and a horizontal distance from the vehicle of r can be defined as being represented by (r, θ, h). Thus, based on the first ground contact point clouds in each determined direction, a data set containing two-dimensional data points in each azimuth angle direction can be formed. In the case of converting and generating two-dimensional data points in this way, each two-dimensional data point generated from the multi-line laser point cloud has a height attribute, and this height attribute is specifically characterized by the ground clearance height h of the corresponding first ground contact point cloud, and this ground clearance height actually corresponds to the ground height of the corresponding first ground contact point cloud. Among them, the ground clearance height of each first ground contact point cloud is specifically calculated using a geometric model constructed based on the vehicle height. Therefore, the two-dimensional data points obtained through this conversion method consider the height of the moving tool itself. In addition to being able to characterize the azimuth of the two-dimensional data points and the distance from the moving tool, it can also effectively characterize the specific position where the end point of the laser scan line in this direction is located. Furthermore, it can effectively assist in obstacle modeling for any obstacle information at a certain height above the ground. In this way, using the solution of the embodiment of the present invention, the obstacles in the blind area and the low objects in the surrounding environment can be effectively processed, avoiding the drawback of misidentifying both the blind area and the low objects as passable areas due to ignoring the height of the moving tool itself. Among them, the specific calculation method of the ground clearance height h of the first ground contact point cloud will be described below in combination with Figure 3 the model effect shown. As Figure 3 shown, since a trapezoidal model 3 is actually defined between the laser end point scanned by the multi-line laser sensor and the vehicle when considering the vehicle's own height, when extending a certain scan line from the position of its first ground contact point cloud to the ground, a triangular area will be defined. Taking scan line A as an example, as Figure 3 shown, by extending scan line A from the position of its first ground contact point cloud to the ground F, a triangular area 4 will be formed. According to the basic data principle, in this triangular area 4, the following mathematical relationship will be formed between the trapezoidal module 3 and the triangular area 4:

[0057] Ground clearance height h / Vehicle height H = Triangular distance t / (Triangular distance t + Horizontal distance r from the ground point cloud to the vehicle);

[0058] Triangular distance t + Horizontal distance r from the ground point cloud to the vehicle = Vehicle height H * cotθ, where θ is the azimuth angle of the scan line.

[0059] Therefore, based on the above mathematical relationship, the ground clearance height at the first ground point cloud position of the corresponding scan line can be determined by the formula h = H - r / cotθ.

[0060] Thus, through the above processing method, the multi-line laser point cloud of any frame scanned by the multi-line laser sensor can be converted into a two-dimensional scan data set corresponding to the corresponding frame, where, Figure 4 Schematically shows the display effect of converting the multi-line laser point cloud of a certain frame into the two-dimensional scan data set corresponding to that frame, as Figure 4 shown. After the above conversion process, the end points of each scan line 5 (the rays emitted from the same source point in the figure) obtained by the multiple laser sensors respectively fall on the ground line 6 (the coil in the figure) or different obstacles. Thus, by determining the first ground point cloud of each scan line and then forming two-dimensional data points based on these ground point clouds, the multi-line laser point cloud is converted into a set of two-dimensional data points defined by the ground clearance height, azimuth angle, and distance of the end points falling on the obstacles, that is, Figure 4 the set of discrete data points 7 in (the set of discrete irregular point-like objects in the figure). It can be Figure 4 seen that these discrete two-dimensional data points 7 have height attributes different from those of the ground line 6.

[0061] After converting the multi-line laser point cloud into a two-dimensional scan data set, the two-dimensional data points in the two-dimensional scan data set can be used to construct and update the grid map. Among them, as a preferred implementation manner, the embodiments of the present invention will construct and update the grid map according to the height attributes and passable attributes of the converted two-dimensional data points, so that the generated grid map can retain more information about low objects and obstacles in the blind area, improving the accuracy and integrity of the generated grid map. For this purpose, in step S11, the embodiments of the present invention will first determine a refresh point set formed by two-dimensional data points assigned with passable attributes from the generated two-dimensional scan data set. Among them, Figure 5 Schematically shows the method for generating a refresh point set based on the two-dimensional scan data set, as Figure 5 shown, and its implementation includes:

[0062] Step S111: According to the distances of the two-dimensional data points in the two-dimensional scan dataset in the current frame and the vehicle radius, determine the two-dimensional data points with distances less than the vehicle radius as valid two-dimensional data points, mark their passable attributes as passable, and add them to the refresh point set;

[0063] Step S112: For the two-dimensional data points in the two-dimensional scan dataset in the current frame with distances greater than or equal to the vehicle radius, select valid two-dimensional data points based on the comparison result of their states with the two-dimensional scan dataset in the historical frame, assign passable attributes to them, and add them to the refresh point set.

[0064] Among them, in step S111, the valid two-dimensional data points refer to the two-dimensional data points that can be used for raster map update. They are relative to the current frame, that is, the two-dimensional data points that can be used for raster map update in the current frame. Due to the movement of the vehicle, some scan points will be surpassed or covered by the vehicle, so they will lose their significance for the construction and update of the raster map in the current frame. Therefore, in the embodiments of the present invention, by determining the valid two-dimensional data points to form the refresh point set, the calculation amount can be reduced and the map generation efficiency can be improved. The vehicle radius refers to the radius of the circle formed with the vehicle width as the diameter and the center point of the vehicle head as the center of the circle, and its value is half of the vehicle width. Since it can be known from empirical data that the area within the vehicle radius is usually a passable area, therefore, in step S111, in the embodiments of the present invention, according to prior knowledge, directly mark the passable attributes of the two-dimensional data points with the horizontal distance r from the vehicle in the current frame less than the vehicle radius as passable, so as to improve the accuracy of the assigned passable attributes of the two-dimensional data points and reduce the calculation amount.

[0065] In step S112, for two-dimensional data points whose distance is not less than the vehicle radius, the embodiments of the present invention will assign passable attributes to the two-dimensional data points of the current frame in combination with the comparison between the state of the two-dimensional data points and the historical state records of the two-dimensional scan data set. Specifically, when assigning passable attributes to the two-dimensional data points of the current frame in combination with the comparison between the state of the two-dimensional data points and the historical state records of the two-dimensional scan data set, three situations will occur: The first situation is that a certain two-dimensional data point in the current frame is also an end point in the historical frame, that is, this two-dimensional data point in the current frame also exists in the two-dimensional scan data set of the historical frame. At this time, it indicates that this two-dimensional data point is an obstacle, so it is added to the refresh point set and its passable attribute is marked as an obstacle; The second situation is that a certain two-dimensional data point in the current frame is not an end point in the historical frame, that is, this two-dimensional data point in the current frame does not exist in the two-dimensional scan data set of the historical frame. At this time, it indicates that this two-dimensional data point may be an uneven road bump or a low object, so it is added to the refresh point set and its passable attribute is temporarily marked as passable; The third situation is that some two-dimensional data points existing in the two-dimensional scan data set of the historical frame do not exist in the two-dimensional data scan set of the current frame, indicating that these two-dimensional data points in the historical frame may have entered the blind area due to the movement of the vehicle in the current frame. Therefore, a judgment will be made on whether these two-dimensional data points are within the vehicle passing height area, that is, whether they are in the blind area. Exemplarily, it can be combined with Figure 3 The geometric model relationship shown is used to determine the ground horizontal distance range of the blind area according to the vehicle height, and the distance r of these two-dimensional data points is compared with the determined blind area distance range to determine whether these two-dimensional data points are within the blind area. If they are within the blind area, these two-dimensional data points are also added to the refresh point set and their passable attribute is marked as an obstacle. It should be noted that the end points in the embodiments of the present invention all refer to the laser end points of the scan line, which are the measurement end points. As described above, they may fall on obstacles or on the ground. In the foregoing steps, the embodiments of the present invention determine the ground contact point cloud in each direction as the reference scan point for generating two-dimensional data points by detecting the end points falling on obstacles. Therefore, the two-dimensional scan points in the two-dimensional scan data set in the current frame are all end points in each direction. Therefore, the above three states can be determined based on the comparison of the two-dimensional data points in the two-dimensional scan data sets in the current frame and the historical frame, and the refresh point set is generated based on the above three states. Thus, the embodiments of the present invention can establish obstacle information at a certain height above any ground, and as long as a part of the obstacle is above the ground and is detected once, this obstacle information can be effectively retained. Therefore, it can fully retain the blind area obstacle information and construct a low object model.

[0066] In a preferred embodiment, when the second situation occurs, that is, when a certain two-dimensional data point in the current frame is not an end point in the historical frame, the gradient model can also be used to calculate the height from the ground of the two-dimensional data point in the historical frame for low object modeling.

[0067] Thus, after the processing of step S11, passable attributes are respectively assigned to a series of valid two-dimensional data points in the refresh point set, and thus a set of points with height and passable attribute marks is obtained. Thus, the embodiment of the present invention can use this point set to generate a grid map. It should be noted that the grid map (or occupancy grid map) is one of the commonly used map types in mobile tools such as robots and autonomous driving vehicles, which can express the position information of obstacles in the map in the form of probability at a certain resolution, providing data input for the autonomous navigation and path planning of robots, autonomous driving vehicles, etc. In the process of constructing a grid map, generally, a lidar is used to scan the environment to obtain point cloud data, and the required grid map is constructed based on the point cloud data. In the embodiment of the present invention, since the two-dimensional data points are generated by converting multi-line lidar point clouds, compared with the two-dimensional point cloud data obtained by a single-line lidar, there is an additional parameter of height from the ground. Therefore, the grid map generated based on the two-dimensional data points can use this height parameter to retain more complete obstacle information, reducing the risk of updating blind area obstacles and low objects into passable areas, and making the generated grid map able to more realistically reflect the occupancy probability of blind areas and low objects.

[0068] In a specific implementation, when generating a grid map using two-dimensional point cloud data, the difference between the embodiment of the present invention and the prior art is that the embodiment of the present invention can update the grid map corresponding to the target area according to the height attribute and passable attribute of the two-dimensional data points in the refresh point set, so as to realize the optimized update of the grid map using two-dimensional point cloud data with height attributes, and protect the obstacles in the blind area and low obstacles from being updated into passable areas. Among them, Figure 5 Schematically shows the method process of updating the grid map corresponding to the target area using the refresh point set in an embodiment of the present invention, as Figure 6 shown, step S12 of the embodiment of the present invention can be specifically implemented as including:

[0069] Step S121: Determine the correspondence between the two-dimensional data points in the refresh point set and the grids in the grid map corresponding to the target area;

[0070] Step S122: Update the occupancy probability and grid height of the corresponding grids according to the passable attribute and height attribute of the two-dimensional data points in the refresh point set.

[0071] In step S121, according to the point cloud coordinates of the multi-line laser point cloud corresponding to the two-dimensional data points, they can be converted into the coordinate positions in the grid map coordinate system according to the center and resolution of the grid map, and the corresponding relationship between the two-dimensional data points and the grids in the grid map can be determined according to the relationship between the converted coordinate positions and the grid positions. Among them, the specific implementation method of converting the two-dimensional data points into the coordinate positions in the grid map coordinate system according to the corresponding point cloud coordinates can refer to the relevant existing technologies and will not be elaborated here.

[0072] In step S122, according to the passable attribute and height attribute of the two-dimensional data points in the refreshed point set, the occupancy probability and grid height of the corresponding grids are updated. Specifically, it can be achieved through Figure 7 the method shown, as Figure 7 shown, and it can be specifically implemented as including:

[0073] Step S1221: When the passable attribute of the two-dimensional data point is an obstacle, increase the occupancy probability of the grid corresponding to the two-dimensional data point and update the grid height of the grid.

[0074] Step S1222: When the passable attribute of the two-dimensional data point is passable, update the occupancy probability and grid height of the corresponding grid according to the grid attribute of the grid corresponding to the two-dimensional data point.

[0075] In step S1221, the occupancy probability of the grid corresponding to the two-dimensional point cloud can be increased by a preset fixed value, and the size of the preset fixed value can be set according to requirements or empirical values. And updating the grid height of the corresponding grid can specifically be adjusting the grid height of the corresponding grid to the ground clearance height of the two-dimensional data point. Since when assigning the passable attribute to the two-dimensional data points, the situations of blind areas being obstacles and low objects being obstacles are fully considered, therefore, when using the two-dimensional data points to update the grid map, directly increasing the occupancy probability of the grid corresponding to the two-dimensional data point with the passable attribute being an obstacle can help construct the low object model and retain the blind area obstacle information to effectively process the blind areas and low objects and improve the authenticity and integrity of the grid map.

[0076] In step S1222, when the passable attribute of the two-dimensional data point is passable, in order to avoid updating the low objects to the passable area, the grid attribute and height attribute of the grid corresponding to the two-dimensional data point can be referred to simultaneously, so as to jointly determine the occupancy probability and grid height of the corresponding grid based on the height attribute and passable attribute of the two-dimensional data point and the height attribute and grid attribute of its corresponding grid. Figure 8 Schematically shows a grid update method when the passable attribute of the two-dimensional data point is passable, as Figure 8 shown, and its implementation includes:

[0077] Step S1222A: When the grid attribute of the grid corresponding to the two-dimensional data point is also passable, reduce the occupancy probability of the grid corresponding to the two-dimensional data point and update the grid height of the grid.

[0078] Step S1222B: When the grid attribute of the grid corresponding to the two-dimensional data point is an obstacle, update the grid height and occupancy probability of the grid according to the comparison result between the grid height of the grid and the ground clearance height of the two-dimensional data point.

[0079] Among them, in step S1222A, the occupancy probability of the corresponding grid can be reduced to a preset fixed value, and the preset fixed value can be set according to requirements or empirical values. The update of the grid height of the grid can specifically be to update the grid height of the corresponding grid to the ground clearance height of the corresponding two-dimensional data point.

[0080] Since the position of the first ground point cloud scanned may change as the vehicle moves, for example, the ground clearance height of the first ground point cloud scanned may be lower or higher. Therefore, when the two-dimensional data point is passable, there is a risk of mis-refreshing low objects or blind area obstacles as passable areas. Therefore, to avoid the problem of mis-refreshing and retain the information of low objects or blind area obstacles as much as possible, in step S1222B, the grid height corresponding to the two-dimensional data point in the passable state is judged to ensure the effectiveness of grid state refreshing. Specifically, when the two-dimensional data point with a higher ground clearance height detected in the historical frame is assigned an obstacle passable attribute, that is, when the grid attribute before update of the grid corresponding to the two-dimensional data point is an obstacle, only when the ground clearance height at this position is detected to be lower again and the assigned passable attribute is passable, it is reasonable to reduce the occupancy probability of this position, that is, it is possible to update the grid at this position to a passable area. Otherwise, if the corresponding grid is updated from an obstacle to passable based on the detected state with a higher ground clearance height, the information of the original low object will be destroyed. Therefore, in the preferred embodiment, specifically, when the passable attribute of the two-dimensional data point is passable and the grid attribute of the corresponding grid is an obstacle, only when the ground clearance height of the two-dimensional data point is less than the grid height, the occupancy probability of the grid corresponding to the two-dimensional data point is reduced by a preset value, and the original grid height of the grid is maintained, that is, the lower grid height is retained without updating the grid height, and only its occupancy probability is reduced; when the ground clearance height of the two-dimensional data point is not less than the grid height, the original grid attribute and grid height of the grid are retained without updating it. Thus, it is ensured that low obstacles are not refreshed as passable areas, protecting the situation where low objects are refreshed in the blind area, and more low object information can be retained in the grid map.

[0081] Thus, the method of the embodiment of the present invention realizes the conversion of multi-line laser point cloud into two-dimensional scan data, and this conversion is based on the constructed geometric model considering the height of the vehicle itself. Therefore, on the one hand, this method avoids the drawback of the decrease in computing efficiency caused by directly using three-dimensional grids. On the other hand, by introducing the geometric model of the vehicle itself, it is possible to fully consider the occupancy of blind areas and low obstacles in the map, avoiding directly determining blind areas and low objects as passable areas due to ignoring the height information of the vehicle, and can effectively model low objects and protect the integrity of blind area information. Moreover, the height attribute is introduced into the two-dimensional scan data generated by the method of the embodiment of the present invention. Therefore, when generating a grid map, the height information of two-dimensional data points and grids is fully considered, which actually realizes the 2.5D expansion of two-dimensional grids, and thus can effectively improve the accuracy and integrity of the generated grid map.

[0082] Figure 9 Schematically shows a grid map generation device according to an embodiment of the present invention. This device can be applied to mobile tools equipped with multi-line laser sensors, such as autonomous / semi-autonomous driving vehicles, unmanned floor sweepers, robots, etc., for generating a grid map based on multi-line laser, so as to ensure a relatively high computing efficiency while making the generated grid map also be able to ensure the integrity of the occupancy of blind areas and low objects as much as possible. As Figure 9 shown, the device includes:

[0083] A point cloud acquisition module 90, configured to acquire multi-line laser point cloud;

[0084] A first point set determination module 91, configured to generate a two-dimensional scan data set according to the multi-line laser point cloud, where the two-dimensional scan data set is a set of two-dimensional data points generated based on scan points meeting a first preset condition, and the two-dimensional data points have height attributes;

[0085] A second point set determination module 92, configured to generate a refreshed point set according to the two-dimensional scan data set, where the refreshed point set is a set formed by two-dimensional data points meeting a second preset condition, and the two-dimensional data points in the refreshed point set also have passable attributes;

[0086] A map update module 93, configured to update the grid map according to the refreshed point set.

[0087] It should be noted that, for the specific implementation processes of each module involved in the grid generation device of the embodiment of the present invention, such as the specific manner in which the first point set determination module generates a two-dimensional scan data set according to the multi-line laser point cloud, the specific implementation manner in which the second point set determination module determines a refreshed point set according to the two-dimensional scan data set, and the specific implementation manner in which the map update module updates the grid map according to the refreshed point set, can all refer to the description in the method part above, and will not be elaborated here.

[0088] Figure 10 Schematically shows a grid map generation device according to another embodiment of the present invention. As shown in the figure, it is implemented to include:

[0089] A memory 1 for storing executable instructions; and

[0090] A processor 2 for executing the executable instructions stored in the memory. When the executable instructions are executed by the processor, the steps of the grid map generation method described in any one of the foregoing embodiments are implemented.

[0091] In specific practice, exemplarily, the above grid map generation device can be applied to autonomous driving vehicles, driverless cleaners, driverless sweepers, robots and other autonomous driving devices or semi-autonomous driving devices to achieve positioning and map construction of these devices, so that these semi-autonomous or driverless tools can generate a grid map based on the three-dimensional point cloud data information scanned by a multi-line laser sensor. While ensuring relatively low computing power requirements, it can protect the information of blind areas and low objects as completely as possible to construct a more accurate and complete grid map for subsequent path planning and driving control, etc.

[0092] Figure 11 Schematically shows a mobile tool according to an embodiment of the present invention. As Figure 11 shown, the mobile tool includes a grid map generation device 70, so that a mobile tool using a multi-line laser sensor can utilize the functions provided by the grid map generation device for map construction, and then perform subsequent processing such as path planning based on the generated grid map. Among them, the grid map generation device can be Figure 10 the grid map generation device shown.

[0093] Optionally, in practical applications, the mobile tool may further include a sensing and recognition module and other planning and control modules, such as a path planning controller, a low-level controller, etc. The functions of the grid map generation device 70 may also be implemented in the sensing and recognition module or the planner, etc. The embodiments of the present invention do not limit this.

[0094] The "mobile tool" referred to in the embodiments of the present invention may be a vehicle of the L0-L5 autonomous driving technology levels defined by the Society of Automotive Engineers International (SAE International) or the Chinese national standard "Automated Driving Classification for Motor Vehicles".

[0095] Exemplarily, the mobile tool may be a vehicle device or a robot device with the following various functions:

[0096] (1) Passenger-carrying function, such as a family car, a bus, etc.;

[0097] (2) Cargo-carrying function, such as a general truck, a van, a semi-trailer truck, an enclosed truck, a tank truck, a flatbed truck, a container truck, a dump truck, a special-structured truck, etc.;

[0098] (3) Tool function, such as a logistics distribution vehicle, an automated guided vehicle (AGV), a patrol vehicle, a crane, a hoist, an excavator, a bulldozer, a forklift, a road roller, a loader, an off-road engineering vehicle, an armored engineering vehicle, a sewage treatment vehicle, a sanitation vehicle, a vacuum cleaner vehicle, a floor washing vehicle, a sprinkler vehicle, a floor sweeping robot, a food delivery robot, a shopping guide robot, a lawn mower, a golf cart, etc.;

[0099] (4) Entertainment function, such as an entertainment vehicle, an automatic driving device in a playground, a segway, etc.;

[0100] (5) Special rescue function, such as a fire truck, an ambulance, an electric power emergency repair vehicle, an engineering emergency rescue vehicle, etc.

[0101] In some embodiments, the embodiments of the present invention provide a non-volatile computer-readable storage medium, in which one or more programs including execution instructions are stored, and the execution instructions can be read and executed by an electronic device (including but not limited to a computer, a server, or a network device, etc.) to execute the grid map generation method of any one of the above embodiments of the present invention.

[0102] In some embodiments, the embodiments of the present invention further provide a computer program product, the computer program product includes a computer program stored on a non-volatile computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, the computer is enabled to execute the grid map generation method of any one of the above embodiments.

[0103] In some embodiments, the embodiments of the present invention further provide an electronic device, which includes: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the grid map generation method of any one of the above embodiments.

[0104] In some embodiments, the embodiments of the present invention further provide a storage medium, on which a computer program is stored, and when the program is executed by a processor, the grid map generation method of any one of the above embodiments is implemented.

[0105] Figure 12It is a schematic hardware structure diagram of a grid map generation device provided by another embodiment of the present invention. The above grid map generation device can be implemented with the structure shown in this figure. For example, Figure 12 as shown, the grid map generation device includes:

[0106] One or more processors 610 and a memory 620. Figure 12 Here, one processor 610 is taken as an example.

[0107] The grid map generation device may further include: an input device 630 and an output device 640.

[0108] The processor 610, the memory 620, the input device 630, and the output device 640 may be connected through a bus or other means. Figure 12 Here, connection through a bus is taken as an example.

[0109] The memory 620, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the grid map generation method in the embodiments of the present invention. The processor 610 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions, and modules stored in the memory 620, that is, implements the grid map generation method in the above method embodiments.

[0110] The memory 620 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the grid map generation method, etc. In addition, the memory 620 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 620 may optionally include a memory remotely provided relative to the processor 610, and these remote memories can be connected to the electronic device through a network. Examples of the above networks include but are not limited to the Internet, enterprise intranets, local area networks, mobile communication networks, and their combinations.

[0111] The input device 630 can receive input digital or character information, and generate signals related to user settings and function control of the image processing device. The output device 640 may include a display device such as a display screen.

[0112] The one or more modules are stored in the memory 620, and when executed by the one or more processors 610, execute the grid map generation method in any of the above method embodiments.

[0113] The above-mentioned product can execute the method provided by the embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. For the technical details not described in detail in this embodiment, reference can be made to the method provided by the embodiment of the present invention.

[0114] The electronic devices in the embodiments of the present invention exist in various forms, including but not limited to:

[0115] (1) Mobile communication devices: These devices are characterized by having mobile communication functions and mainly aim to provide voice and data communication. Such terminals include: smart phones (such as iPhone), multimedia phones, functional phones, and low-end phones, etc.

[0116] (2) Ultra-mobile personal computer devices: These devices belong to the category of personal computers, have computing and processing functions, and generally also have the characteristic of mobile Internet access. Such terminals include: PDA, MID, and UMPC devices, etc., such as iPad.

[0117] (3) Portable entertainment devices: These devices can display and play multimedia content. Such devices include: audio and video players (such as iPod), handheld game consoles, e-books, and smart toys and portable vehicle navigation devices.

[0118] (4) Servers: Devices that provide computing services. The composition of a server includes a processor, a hard disk, a memory, a system bus, etc. A server is similar to a general computer architecture, but due to the need to provide highly reliable services, it has higher requirements in terms of processing power, stability, reliability, security, scalability, and manageability.

[0119] (5) Other electronic devices with data interaction functions.

[0120] The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0121] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A grid map generation method, characterized in that, The method includes: Obtaining multi-line laser point cloud, and generating a two-dimensional scan data set according to the multi-line laser point cloud, wherein the two-dimensional scan data set is a set of two-dimensional data points generated based on scan points meeting a first preset condition, and the two-dimensional data points have height attributes; Generating a refreshed point set according to the two-dimensional scan data set, wherein the refreshed point set is a set formed by valid two-dimensional data points assigned with passable attributes, and the valid two-dimensional data points in the refreshed point set have both height attributes and passable attributes; Updating a grid map according to the refreshed point set, specifically updating the grid map based on the height attributes and passable attributes of the two-dimensional data points in the refreshed point set; The generating a refreshed point set according to the two-dimensional scan data set includes: According to the distance between the two-dimensional data points in the two-dimensional scan data set in the current frame and the vehicle radius, determining the two-dimensional data points with a distance less than the vehicle radius as valid two-dimensional data points, marking their passable attributes as passable, and adding them to the refreshed point set; For the two-dimensional data points in the two-dimensional scan data set in the current frame with a distance greater than or equal to the vehicle radius, valid two-dimensional data points are selected based on the comparison result of their states with the two-dimensional scan data set in the historical frame, and passable attributes are assigned to the selected valid two-dimensional data points and they are added to the refreshed point set.

2. The method according to claim 1, wherein The scan points meeting the first preset condition are the first ground-contact point clouds in each direction of the multi-line laser point cloud. The generating a two-dimensional scan data set according to the multi-line laser point cloud includes: Sorting the multi-line laser point cloud according to the azimuth angle, and sorting the multi-line laser point cloud in each azimuth angle direction according to the distance; Determining the first ground-contact point cloud in each azimuth angle direction of the multi-line laser point cloud; Generating a two-dimensional scan data set according to the determined first ground-contact point clouds in each azimuth angle direction.

3. The method according to claim 2, characterized in that The determining the first ground-contact point cloud in each azimuth angle direction of the multi-line laser point cloud includes: Constructing a geometric model based on the height of the mobile tool itself, and calculating the ground clearance of each point cloud in each azimuth angle direction of the multi-line laser point cloud according to the constructed geometric model; Determining the first ground-contact point cloud in the corresponding azimuth angle direction according to the ground clearance of each point cloud in each azimuth angle direction of the multi-line laser point cloud and the preset ground parameters.

4. The method according to claim 2, characterized in that, The generating a two-dimensional scan data set according to the determined first ground-contact point clouds in each azimuth angle direction includes: Generating two-dimensional data points corresponding to each first ground-contact point cloud according to the ground clearance, azimuth angle and distance of each first ground-contact point cloud, wherein the two-dimensional data points are jointly defined by the ground clearance, azimuth angle and distance, and the height attribute of the two-dimensional data points is characterized by the ground clearance; Forming the two-dimensional data points corresponding to the first ground-contact point clouds in each azimuth angle direction into a two-dimensional scan data set.

5. The method according to claim 1, wherein For the two-dimensional data points in the two-dimensional scan data set in the current frame with a distance greater than or equal to the vehicle radius, selecting valid two-dimensional data points based on the two-dimensional scan data set in the historical frame, assigning passable attributes to them, and adding them to the refreshed point set, includes: According to the matching result between the two-dimensional data points in the current frame and the two-dimensional data points in the two-dimensional scan data set in the historical frame, determine whether the two-dimensional data point is an end point in the historical frame. If the two-dimensional data point is an end point in the historical frame, determine the two-dimensional data point in the current frame as a valid two-dimensional data point, mark its passable attribute as an obstacle, and add it to the refresh point set; if the two-dimensional data point is not an end point in the historical frame, determine the two-dimensional data point in the current frame as a valid two-dimensional data point, mark its passable attribute as passable, and add it to the refresh point set; Extract the two-dimensional data points that are inconsistent with the two-dimensional data points in the current frame from the two-dimensional scan data set of the historical frame, and determine whether these inconsistent two-dimensional data points are within the vehicle passing height range. Determine the two-dimensional data points that are inconsistent with the two-dimensional scan points in the two-dimensional scan data set of the current frame and are within the vehicle passing height range as valid two-dimensional data points, mark their passable attribute as an obstacle, and add them to the refresh point set.

6. The method according to claim 5, characterized in that, Further includes: When the two-dimensional data point is not an end point in the historical frame, use the gradient model to calculate and record the ground clearance of the two-dimensional data point that is not an end point.

7. The method according to any one of claims 1 to 6, characterized in that The updating of the grid map according to the refresh point set includes: Determine the correspondence between the two-dimensional data points in the refresh point set and the grids in the grid map corresponding to the target area; Update the occupancy probability and grid height of the corresponding grids according to the passable attributes of the two-dimensional data points in the refresh point set.

8. The method according to claim 7, wherein The updating of the occupancy probability and grid height of the corresponding grids according to the passable attributes of the two-dimensional data points in the refresh point set includes; When the passable attribute of the two-dimensional data point is an obstacle, increase the occupancy probability of the grid corresponding to the two-dimensional data point, and update the grid height of the grid; When the passable attribute of the two-dimensional data point is passable, update the occupancy probability and grid height of the corresponding grid according to the grid attribute and grid height of the grid corresponding to the two-dimensional data point.

9. The method according to claim 8, wherein When the passable attribute of the two-dimensional data point is passable, updating the occupancy probability and grid height of the corresponding grid according to the grid attribute and grid height of the grid corresponding to the two-dimensional data point includes: When the grid attribute of the grid corresponding to the two-dimensional data point is also passable, decrease the occupancy probability of the grid corresponding to the two-dimensional data point, and update the grid height of the grid; When the grid attribute of the grid corresponding to the two-dimensional data point is an obstacle, according to the comparison result between the grid height of the grid and the ground clearance of the two-dimensional data point, when the ground clearance of the two-dimensional data point is less than the grid height of the corresponding grid, decrease the occupancy probability of the grid.

10. A grid map generation device, characterized in that, The device includes: A point cloud acquisition module for acquiring multi-line laser point cloud; A first point set determination module for generating a two-dimensional scan data set according to the multi-line laser point cloud, wherein, the two-dimensional scan data set is a set of two-dimensional data points generated based on scan points that meet the first preset condition, and the two-dimensional data points have height attributes; A second point set determination module, configured to generate a refreshed point set according to the two-dimensional scan data set, where the refreshed point set is a set formed by valid two-dimensional data points assigned with a passable attribute, and the two-dimensional data points in the refreshed point set have both a height attribute and a passable attribute. The generating the refreshed point set according to the two-dimensional scan data set includes determining, according to the distance between the two-dimensional data points in the two-dimensional scan data set in the current frame and the vehicle radius, the two-dimensional data points with a distance less than the vehicle radius as valid two-dimensional data points, marking their passable attributes as passable, and adding them to the refreshed point set; for the two-dimensional data points in the two-dimensional scan data set in the current frame with a distance greater than or equal to the vehicle radius, valid two-dimensional data points are selected based on the comparison result of their states with the two-dimensional scan data set in the historical frame, and passable attributes are assigned to the selected valid two-dimensional data points and they are added to the refreshed point set. A map update module, configured to update the grid map according to the refreshed point set, specifically, updating the grid map based on the height attribute and the passable attribute of the two-dimensional data points in the refreshed point set.

11. A grid map generation device, characterized in that, Comprising: A memory, configured to store executable instructions; And A processor, configured to execute the executable instructions stored in the memory, and the executable instructions, when executed by the processor, implement the steps of the method according to any one of claims 1 to 9.

12. Mobile tool, characterized in that, The mobile tool includes: the grid map generation device according to claim 11.

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

14. A computer program product, the computer program product includes a computer program stored on a non-volatile computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a computer, the computer is caused to execute the method according to any one of claims 1 to 9.

Images