Road fence detection method, device, equipment and medium

Abstract

The embodiment of the invention discloses a road fence detection method and device, equipment and a medium. The method comprises the following steps: acquiring an obstacle list of a current vehicle, and determining a seed obstacle and an obstacle to be detected from the obstacle list; adding the seed obstacles into a fence obstacle list, and determining the number of the fence obstacles in the fence obstacle list; determining a current obstacle to be detected from the obstacles to be detected, and selecting a target fence detection strategy from preset candidate fence detection strategies according to the number of the obstacles; according to the target fence detection strategy, fence detection is carried out on the current to-be-detected obstacle, whether the current to-be-detected obstacle is a target obstacle is judged, and if yes, the target obstacle is added to a fence obstacle list; and after fence detection of all the to-be-detected obstacles is completed, determining the fence on the road where the current vehicle is located according to the fence obstacles in the fence obstacle list. And the detection accuracy of the fence is improved.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of autonomous driving technology, and in particular to a road barrier detection method, device, equipment and medium. Background Art

[0002] There are two main methods for obstacle detection based on LiDAR point clouds. The first is a deep learning-based target detection system that uses a convolutional neural network to regress information such as the target's center point, size, orientation, and category. The second is based on geometric rules, using ground plane fitting, point cloud clustering, and other methods to construct target obstacles. Each method has its own advantages and disadvantages, so most current solutions are a combination of the two methods.

[0003] However, the rule-based obstacle segmentation algorithm has many limitations because it is completely based on some rules and experience settings, and the algorithm itself cannot understand the actual scene information.

[0004] A common limitation is that the point cloud input to the algorithm needs to be filtered by the map. According to the road information provided by the map, the points inside the road are taken as the processing objects. First, they are preprocessed and ground detected to extract non-ground points. After clustering, all non-ground points are output as obstacles. If some background points are divided into the inside of the road, then an obstacle that is not a foreground target is output, which affects driving. For example, a fence (or railing) that is not set on the road boundary may be judged as an obstacle that is not a foreground target, which affects vehicle driving. Therefore, it is crucial to improve the accuracy of fence detection. Summary of the invention

[0005] The present invention provides a road fence detection method, device, equipment and medium to improve the accuracy of detecting fences on the road.

[0006] According to one aspect of the present invention, there is provided a road barrier detection method, comprising:

[0007] Obtaining an obstacle list of the current vehicle, and determining a seed obstacle and an obstacle to be detected from the obstacle list;

[0008] Adding the seed obstacle to a barrier obstacle list, and determining the number of barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include seed obstacles and / or target obstacles;

[0009] Determine the current obstacle to be detected from the obstacles to be detected, and select a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles;

[0010] According to the target barrier detection strategy, barrier detection is performed on the current obstacle to be detected to determine whether the current obstacle to be detected is the target obstacle, and if so, the target obstacle is added to the barrier obstacle list;

[0011] After the barriers of the obstacles to be detected are all detected, the barriers on the road where the current vehicle is located are determined according to the barrier obstacles in the barrier obstacle list.

[0012] According to another aspect of the present invention, there is provided a road barrier detection device, comprising:

[0013] An obstacle determination module is used to obtain an obstacle list of the current vehicle and determine a seed obstacle and an obstacle to be detected from the obstacle list;

[0014] An obstacle quantity determination module, used for adding the seed obstacle to the barrier obstacle list, and determining the obstacle quantity of the barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include seed obstacles and / or target obstacles;

[0015] A detection strategy determination module, used to determine the current obstacle to be detected from the obstacles to be detected, and select a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles;

[0016] A barrier detection module, configured to perform barrier detection on the obstacle to be detected according to the target barrier detection strategy, determine whether the obstacle to be detected is the target obstacle, and if so, add the target obstacle to the barrier obstacle list;

[0017] The barrier determination module is used to determine the barriers on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list after the barrier detection of the obstacles to be detected is completed.

[0018] According to another aspect of the present invention, there is provided an electronic device, comprising:

[0019] one or more processors;

[0020] A memory for storing one or more programs;

[0021] When one or more programs are executed by one or more processors, the one or more processors can execute any road barrier detection method provided by the embodiments of the present invention.

[0022] According to another aspect of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement any road barrier detection method provided by the embodiments of the present invention when executed.

[0023] The embodiment of the present invention provides a road barrier detection scheme, which obtains the obstacle list of the current vehicle, determines the seed obstacle and the obstacle to be detected from the obstacle list; adds the seed obstacle to the barrier obstacle list, and determines the obstacle quantity of the barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include the seed obstacle and / or the target obstacle; determines the current obstacle to be detected from the obstacles to be detected, and selects the target barrier detection strategy from the preset candidate barrier detection strategies according to the obstacle quantity; performs barrier detection on the current obstacle to be detected according to the target barrier detection strategy, determines whether the current obstacle to be detected is the target obstacle, and if so, adds the target obstacle to the barrier obstacle list; after the barrier detection of all obstacles to be detected is completed, determines the barrier on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list. The above scheme improves the accuracy of barrier detection by determining the barrier obstacle according to the seed obstacle and the obstacle to be detected, and then determining the barrier on the road according to the barrier obstacle.

[0024] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1A is a flow chart of a road barrier detection method provided by Embodiment 1 of the present invention;

[0027] Figure 1B This is a schematic diagram of a laser radar scanning a fence provided by the first embodiment of the present invention;

[0028] Figure 2A is a flow chart of a road barrier detection method provided by Embodiment 2 of the present invention;

[0029] Figure 2BIt is a schematic diagram of determining whether the obstacle to be detected is a target obstacle based on an angle detection strategy provided by the second embodiment of the present invention;

[0030] Figure 3 is a schematic diagram of the structure of a road barrier detection device provided in Embodiment 3 of the present invention;

[0031] Figure 4 It is a structural schematic diagram of an electronic device for implementing a road barrier detection method provided by a fourth embodiment of the present invention. DETAILED DESCRIPTION

[0032] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0033] In the prior art, the road information provided by the map is the internal area of ​​the lane line. Then there is a problem in this scenario. If the fence is outside the road area, it will not be output. However, since the lane line of the non-motorized vehicle lane is not considered a road boundary, the fence on one side is inside the map. Then the geometric segmentation algorithm will output the fence target. For the downstream, this fence target is a false detection. Therefore, a method is needed to detect this fence target and filter it out. Therefore, an embodiment of the present invention proposes a method for detecting fences on non-boundaries of roads.

[0034] Embodiment 1

[0035] Figure 1A This is a flow chart of a road barrier detection method provided in Embodiment 1 of the present invention. This embodiment is applicable to the case of detecting barriers on a road. The method can be executed by a road barrier detection device, which can be implemented in software and / or hardware and can be configured in an electronic device that carries a road barrier detection function.

[0036] See also Figure 1A The road barrier detection method shown includes:

[0037] S110: Obtain an obstacle list of the current vehicle, and determine a seed obstacle and an obstacle to be detected from the obstacle list.

[0038] The current vehicle refers to the vehicle that provides the obstacle list. The obstacle list refers to the foreground target list obtained by processing the point cloud, that is, the obstacle list determined by the point cloud data on the road collected by the laser radar on the current vehicle. The seed obstacle refers to the starting obstacle used to determine the fence on the road. The obstacle to be detected refers to the obstacle that needs to be detected by the fence.

[0039] In an optional embodiment, determining seed obstacles and obstacles to be detected from the obstacle list includes: determining obstacles to be screened from the candidate obstacles according to obstacle attribute parameters of each candidate obstacle in the obstacle list and a preset obstacle screening strategy; sorting each obstacle to be screened according to the obstacle center point of each obstacle to be screened to obtain an obstacle sorting result; determining seed obstacles from the obstacles to be screened according to the obstacle attribute parameters of each obstacle to be screened and a preset seed determination strategy; and determining obstacles to be detected according to the seed obstacles and the obstacle sorting result.

[0040] The candidate obstacles refer to obstacles stored in the obstacle list. The embodiment of the present invention does not impose any limitation on the method of determining each candidate obstacle in the obstacle list, which can be set by technicians based on experience. The obstacle screening strategy refers to a strategy that can be used to screen candidate obstacles.

[0041] Among them, the obstacle attribute parameters refer to the basic attribute data of the candidate obstacles. Exemplarily, the obstacle attribute parameters may include: the obstacle center point, obstacle size, obstacle direction span, the minimum height of the obstacle from the ground min_height and the point cloud contained in the obstacle. Among them, the obstacle center point may include center_x, center_y and center_z, that is, the center value of the candidate obstacle in the three directions of X (front), Y (left) and Z (top). The obstacle size may include the width of the obstacle, the height of the obstacle and the length of the obstacle. The direction span of the obstacle refers to the difference between the minimum point and the maximum point of the envelope in the three directions of XYZ, that is, the direction span of the obstacle may include the X-direction span d_x, the Y-direction span d_y and the Z-direction span d_z.

[0042] It should be noted that the obstacle list includes candidate obstacles and obstacle attribute parameters corresponding to the candidate obstacles.

[0043] The obstacles to be screened refer to obstacles screened from candidate obstacles based on the obstacle screening strategy. The obstacle sorting result refers to the result of sorting the obstacles to be screened according to the obstacle center points of each obstacle to be screened. Exemplarily, the forward center value (i.e., the X-direction center value center_x) of the obstacle center point of each obstacle to be screened is determined; according to the forward center value center_x of each obstacle to be screened, the obstacles to be screened are sorted from small to large to obtain the obstacle sorting result.

[0044] Among them, the seed determination strategy can be used to determine the seed obstacles from the obstacles to be screened.

[0045] For example, the candidate obstacles are initially screened to obtain the obstacles to be screened. As Figure 1B shown in the schematic diagram of the lidar scanning the railing. Since the railing is generally a narrow and long target parallel to the driving direction, its point cloud also satisfies this feature. However, due to the limited resolution of the lidar, when the distance becomes farther, the point cloud scanned by the railing will become discontinuous and turn into small adjacent vertical strip-shaped point clouds. This will result in the obstacle to be screened being in a long strip shape at a close distance and a small target with both small length and width at a far distance. Therefore, for the candidate obstacles, the present invention sets an obstacle screening strategy with the following conditions. The candidate obstacles that meet the following conditions are the obstacles to be screened:

[0046] center_x < railings_max_dis;

[0047] height < railings_max_height;

[0048] width < railings_max_width or (length > railings_min_length and

[0049] length / width > railings_min_ratio)

[0050] Among them, railings_max_dis represents the X-direction center value threshold; railings_max_height represents the railing height threshold; railings_max_width represents the railing width threshold; railings_min_length represents the railing length threshold; railings_min_ratio represents the railing length-width ratio threshold. In the embodiments of the present invention, the sizes of the X-direction center value threshold, the railing height threshold, the railing width threshold, the railing length threshold, and the railing length-width ratio threshold are not limited in any way. They can be set by those skilled in the art according to experience or determined repeatedly through a large number of experiments.

[0051] The obstacle screening strategy in the embodiments of the present invention may include: the X-direction center value of the candidate obstacle is less than the X-direction center value threshold, the height of the candidate obstacle is less than the railing height threshold, the width of the candidate obstacle is less than the railing width threshold, the length of the candidate obstacle is greater than the railing length threshold, and the length-width ratio of the candidate obstacle is greater than the railing length-width ratio threshold, etc.

[0052] Continuing with the above example, the above-mentioned obstacles to be screened are sorted. According to the size of the X-direction (i.e., forward) center value, the obstacles to be screened are sorted from small to large to obtain the obstacle sorting result. The advantage is that after sorting, it helps to search for obstacles more quickly in the subsequent process. By traversing in order, the remaining obstacles can be merged in turn, saving computing resources.

[0053] Continuing with the above example, we traverse all obstacles to be screened and find obstacles to be screened that are close and can be used as seed obstacles. The obstacle to be screened that satisfies the following seed determination strategy is considered a seed obstacle:

[0054] length>seed_length;

[0055] width <seed_width;

[0056] length / width>seed_ratio;

[0057] dx>dy;

[0058] min_height>seed_min_height;

[0059] Wherein, seed_length represents the seed length threshold; seed_width represents the seed width threshold; seed_ratio represents the seed aspect ratio threshold; seed_min_height represents the seed minimum height threshold. The embodiment of the present invention does not impose any limitation on the size of the seed length threshold, the seed width threshold, the seed aspect ratio threshold, and the seed minimum height threshold, which can be set by technicians based on experience or repeatedly determined through a large number of experiments.

[0060] It should be noted that the seed length threshold and the fence length threshold in the embodiment of the present invention may be the same or different. Exemplarily, the seed length threshold may be smaller than the fence length threshold. The seed width threshold and the fence width threshold may be the same or different. Exemplarily, the seed width threshold may be smaller than the fence width threshold. The seed aspect ratio threshold and the fence aspect ratio threshold may be the same or different. Exemplarily, the seed aspect ratio threshold may be smaller than the fence aspect ratio threshold.

[0061] The seed determination strategy in the embodiment of the present invention may include: the length of the obstacle to be screened is greater than the seed length threshold, the width of the obstacle to be screened is less than the seed width threshold, the length-to-width ratio of the obstacle to be screened is greater than the seed length-to-width ratio threshold, the X-direction span of the obstacle to be screened is greater than the Y-direction span, and the height of the lowest point of the obstacle to be screened from the ground is greater than the seed minimum height threshold, etc.

[0062] Continuing with the previous example, with the seed obstacle as a reference, the obstacle sorting result is traversed forward, and the obstacle to be detected is determined from the obstacles to be screened other than the seed obstacle; or, with the seed obstacle as a reference, the obstacle sorting result is traversed backward, and the obstacle to be detected is determined from the obstacles to be screened other than the seed obstacle. The embodiment of the present invention is described by taking the forward traversal as an example.

[0063] It can be understood that by introducing the obstacle screening strategy and the seed determination strategy, the seed obstacles and obstacles to be detected are determined from the candidate obstacles in the obstacle list, thereby improving the accuracy of the determined seed obstacles and obstacles to be detected.

[0064] S120: Add the seed obstacle to the barrier obstacle list, and determine the number of barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include the seed obstacle and / or the target obstacle.

[0065] The barrier obstacle list can be used to store barrier obstacles. The barrier obstacle can be used to determine the barrier on the road where the current vehicle is located. The number of obstacles refers to the number of barrier obstacles in the barrier obstacle list. The target obstacle refers to the obstacle selected from the obstacles to be detected for determining the barrier.

[0066] S130: Determine the current obstacle to be detected from the obstacles to be detected, and select a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles.

[0067] The candidate barrier detection strategy refers to a pre-built detection strategy for determining whether the obstacle to be detected is a target obstacle. The target barrier detection strategy refers to a candidate barrier detection strategy for determining whether the current obstacle to be detected is a target obstacle. The current obstacle to be detected refers to the obstacle to be detected that is currently undergoing barrier detection.

[0068] It should be noted that the current obstacle to be detected is determined from the obstacles to be detected by: according to the obstacle sorting result and the obstacles to be detected, taking the seed obstacle as the reference, the current obstacles to be detected are determined one by one in the direction away from the seed obstacle. At the same time, it should be noted that some obstacles to be detected that have been fenced should be marked to prevent repeated traversal.

[0069] S140: Perform barrier detection on the current obstacle to be detected according to the target barrier detection strategy to determine whether the current obstacle to be detected is a target obstacle. If so, add the target obstacle to the barrier obstacle list.

[0070] Exemplarily, if it is determined that the current obstacle to be detected is not a target obstacle, it is prohibited to add the current obstacle to be detected to the fence obstacle list.

[0071] S150: After all obstacles to be detected have completed barrier detection, determine barriers on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list.

[0072] Exemplarily, if all obstacles to be detected have been subjected to barrier detection, then the barrier obstacles in the barrier obstacle list are merged to obtain the barriers on the road where the current vehicle is located.

[0073] The embodiment of the present invention provides a road barrier detection scheme, which obtains the obstacle list of the current vehicle, determines the seed obstacle and the obstacle to be detected from the obstacle list; adds the seed obstacle to the barrier obstacle list, and determines the obstacle quantity of the barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include the seed obstacle and / or the target obstacle; determines the current obstacle to be detected from the obstacles to be detected, and selects the target barrier detection strategy from the preset candidate barrier detection strategies according to the obstacle quantity; performs barrier detection on the current obstacle to be detected according to the target barrier detection strategy, determines whether the current obstacle to be detected is the target obstacle, and if so, adds the target obstacle to the barrier obstacle list; after the barrier detection of all obstacles to be detected is completed, determines the barrier on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list. The above scheme improves the accuracy of barrier detection by determining the barrier obstacle according to the seed obstacle and the obstacle to be detected, and then determining the barrier on the road according to the barrier obstacle.

[0074] Embodiment 2

[0075] Figure 2A This is a flow chart of a road barrier detection method provided in the second embodiment of the present invention. Based on the above embodiments, this embodiment further refines the operation of "selecting a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles" into "if the number of obstacles is 1, the target barrier detection strategy is a straight line detection strategy; if the number of obstacles is 2, the target barrier detection strategy is an angle detection strategy; if the number of obstacles exceeds 2, the target barrier detection strategy includes an angle detection strategy and / or a line fitting detection strategy" to improve the determination mechanism of the target barrier detection strategy. It should be noted that for the parts not described in detail in the embodiments of the present invention, reference can be made to the descriptions of other embodiments.

[0076] See also Figure 2A The road barrier detection method shown includes:

[0077] S210: Obtain an obstacle list of the current vehicle, and determine a seed obstacle and an obstacle to be detected from the obstacle list.

[0078] S220: Add the seed obstacle to the barrier obstacle list, and determine the number of barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include the seed obstacle and / or the target obstacle.

[0079] S230: Determine a current obstacle to be detected from obstacles to be detected.

[0080] S240: If the number of obstacles is 1, the target fence detection strategy is a straight line detection strategy.

[0081] Among them, the straight line detection strategy refers to a strategy for judging whether the obstacle to be detected is a target obstacle based on the seed obstacle line.

[0082] In an optional embodiment, if the target fence detection strategy is a straight line detection strategy, fence detection is performed on the current obstacle to be detected according to the target fence detection strategy to determine whether the current obstacle to be detected is a target obstacle, including: obtaining the seed center point of the seed obstacle, and dividing the seed coordinate points in the seed obstacle into a first coordinate set and a second coordinate set according to the seed center point; determining the first coordinate mean and the second coordinate mean in the first coordinate set and the second coordinate set respectively, and determining the seed obstacle line according to the first coordinate mean and the second coordinate mean; determining the seed distance between the current obstacle to be detected and the seed obstacle line, and determining whether the current obstacle to be detected is a target obstacle according to the seed distance and a preset seed distance threshold.

[0083] Among them, the seed center point refers to the obstacle center point of the seed obstacle. The seed coordinate point refers to the coordinate point of the point cloud in the seed obstacle. Exemplarily, the seed coordinate point is divided into two sets, namely the first coordinate set and the second coordinate set, with the X-direction center value in the seed center point as the boundary. For example, the first coordinate set may include some seed coordinate points before the X-direction center value, and correspondingly, the second coordinate set may include some seed coordinate points after the X-direction center value; or, the first coordinate set may include some seed coordinate points after the X-direction center value, and correspondingly, the second coordinate set may include some seed coordinate points before the X-direction center value.

[0084] The first coordinate mean refers to the mean of various sub-coordinate points in the first coordinate set. The second coordinate mean refers to the mean of various sub-coordinate points in the second coordinate set. The seed obstacle line refers to a straight line determined based on the first coordinate mean and the second coordinate mean, that is, the seed obstacle line is the connecting line between the first coordinate mean and the second coordinate mean.

[0085] The seed distance refers to the distance between the current obstacle to be detected and the seed obstacle line. For example, the distance between the obstacle center point of the current obstacle to be detected and the seed obstacle line is used as the seed distance. The embodiment of the present invention does not impose any limitation on the size of the preset seed distance threshold, which can be set by technicians based on experience or repeatedly determined through a large number of experiments.

[0086] Exemplarily, judging whether the current obstacle to be detected is a target obstacle based on the seed distance and the preset seed distance threshold includes: if the seed distance is less than the preset seed distance threshold, determining that the current obstacle to be detected is a target obstacle; if the seed distance is greater than or equal to the preset seed distance threshold, prohibiting the current obstacle to be detected from being a target obstacle.

[0087] For example, if there is only one fence obstacle in the fence obstacle list, that is, only the seed obstacle S, then it is necessary to determine whether the current obstacle A to be detected is the target obstacle based on the trend of the point cloud in the seed obstacle. First, calculate the two front and rear centers of the seed obstacle S (i.e., the first coordinate mean and the second coordinate mean) to simply calculate the slope and intercept of the seed obstacle line corresponding to the seed obstacle S. The front and rear centers are calculated by taking the X-axis center value of the seed obstacle as the boundary, and calculating the first coordinate mean of all obstacle points in front of the X-axis center value and the second coordinate mean of all obstacle points behind the X-axis center value. An approximate seed obstacle line can be calculated based on the first coordinate mean and the second coordinate mean, and then the seed distance from the obstacle center point of the current obstacle A to be detected to the seed obstacle line is calculated. If the seed distance is less than the preset seed distance threshold first_obj_thr, it is considered that the current obstacle A to be detected is on the same straight line as the seed obstacle S, and the current obstacle A to be detected is the target obstacle and is put into the fence obstacle list; otherwise, jump out of the current loop and traverse the next obstacle to be detected.

[0088] It can be understood that by determining whether the current obstacle to be detected is a target obstacle based on the seed distance between the seed obstacle line and the current obstacle to be detected and the preset seed distance threshold, the accuracy of determining the target obstacle is improved.

[0089] S250: If the number of obstacles is 2, the target fence detection strategy is an angle detection strategy.

[0090] The angle detection strategy refers to a strategy for determining whether the obstacle to be detected is a target obstacle based on the obstacle angle.

[0091] In an optional embodiment, if the target fence detection strategy is an angle detection strategy, fence detection is performed on the current obstacle to be detected according to the target fence detection strategy to determine whether the current obstacle to be detected is a target obstacle, including: determining a first obstacle line between the seed obstacle and the current obstacle to be detected, and determining a second obstacle line between the seed obstacle and the target obstacle; determining an obstacle angle between the first obstacle line and the second obstacle line, and determining whether the current obstacle to be detected is a target obstacle based on the obstacle angle and a preset angle threshold.

[0092] The first obstacle line refers to a straight line determined based on the seed obstacle and the obstacle to be detected currently. Exemplarily, the first obstacle line is determined based on the obstacle center point of the seed obstacle and the obstacle center point of the obstacle to be detected currently, that is, the first obstacle line is a connecting line between the obstacle center point of the seed obstacle and the obstacle center point of the obstacle to be detected currently.

[0093] The second obstacle line refers to a straight line determined based on the seed obstacle and the target obstacle in the fence obstacle list. Exemplarily, the second obstacle line is determined based on the obstacle center point of the seed obstacle and the obstacle center point of the target obstacle, that is, the second obstacle line is a connecting line between the obstacle center point of the seed obstacle and the obstacle center point of the target obstacle.

[0094] The obstacle angle refers to the angle between two lines. The embodiment of the present invention does not impose any limitation on the value of the preset angle threshold, which can be set by technicians based on experience or repeatedly determined through a large number of experiments.

[0095] Exemplarily, judging whether the obstacle to be detected is a target obstacle based on the obstacle angle and the preset angle threshold includes: if the obstacle angle is less than the preset angle threshold, determining that the obstacle to be detected is a target obstacle; if the obstacle angle is greater than or equal to the preset angle threshold, prohibiting the obstacle to be detected from being a target obstacle.

[0096] For example, if there are already two fence obstacles in the fence obstacle list, then it is necessary to calculate the obstacle angle between the current obstacle A to be detected and the obstacle center point of the existing fence obstacle to determine whether the current obstacle A to be detected is the target obstacle. The specific judgment method is as follows Figure 2BThe long object at the bottom is the seed obstacle S0, the small object in the middle is the target obstacle S1 previously classified, and the top object is the current obstacle A to be detected. The obstacle center point connecting the seed obstacle S0 and the current obstacle A to be detected (i.e., the first obstacle line) is calculated, and the obstacle center point connecting the target obstacle S1 and the seed obstacle S0 (i.e., the second obstacle line) is calculated to determine the obstacle angle theta between the first obstacle line and the second obstacle line; if the obstacle angle is less than the preset angle threshold ranlings_thrta_thr, then the current obstacle A to be detected is the target obstacle, as shown in Figure 1. Figure 2B In the case shown in the middle left part, the obstacle A to be detected is put into the fence obstacle list; if not, the obstacle to be detected is not the target obstacle, such as Figure 2B In the case shown in the middle right part, the current loop is jumped out and the next obstacle to be detected is traversed.

[0097] It can be understood that by determining whether the current obstacle to be detected is a target obstacle based on the obstacle angle and the preset angle threshold, the accuracy of determining the target obstacle is improved.

[0098] S260: If the number of obstacles exceeds 2, the target fence detection strategy includes an angle detection strategy and / or a line fitting detection strategy.

[0099] Among them, the line fitting detection strategy is a strategy for judging whether the obstacle to be detected is a target obstacle based on the obstacle fitting line.

[0100] In an optional embodiment, if the target fence detection strategy includes an angle detection strategy and / or a line fitting detection strategy, fence detection is performed on the current obstacle to be detected according to the target fence detection strategy to determine whether the current obstacle to be detected is a target obstacle, including: determining the center distance between the obstacle center point of each fence obstacle in the fence obstacle list and the obstacle center point of the current obstacle to be detected, and selecting a first reference obstacle and a second reference obstacle from the fence obstacles according to the center distance; determining a first reference line between the first reference obstacle and the current obstacle to be detected, and determining a second reference line between the first reference obstacle and the second reference obstacle; determining an obstacle angle between the first reference line and the second reference line, and determining whether the current obstacle to be detected is a target obstacle according to the obstacle angle and a preset angle threshold; if not, fence detection is performed on the current obstacle to be detected according to the line fitting detection strategy to determine whether the current obstacle to be detected is a target obstacle.

[0101] The center distance refers to the distance between the obstacle center point of the barrier obstacle and the obstacle center point of the obstacle to be detected. The second reference obstacle refers to the barrier obstacle closest to the obstacle to be detected, that is, the barrier obstacle with the smallest center distance. The first reference obstacle refers to the barrier obstacles other than the second reference obstacle that are closest to the obstacle to be detected, that is, the barrier obstacle with the second smallest center distance.

[0102] The first reference line refers to a straight line determined based on the first reference obstacle and the obstacle to be detected. Exemplarily, the first reference line is determined based on the obstacle center point of the first reference obstacle and the obstacle center point of the obstacle to be detected, that is, the first reference line is a connecting line between the obstacle center point of the first reference obstacle and the obstacle center point of the obstacle to be detected.

[0103] The second reference line refers to a straight line determined based on the first reference obstacle and the second reference obstacle. Exemplarily, the second reference line is determined based on the obstacle center point of the first reference obstacle and the obstacle center point of the second reference obstacle, that is, the second reference line is a connecting line between the obstacle center point of the first reference obstacle and the obstacle center point of the second reference obstacle.

[0104] Exemplarily, after determining whether the obstacle to be detected is a target obstacle based on the obstacle angle and the preset angle threshold, the method further includes: if so, determining that the obstacle to be detected is a target obstacle.

[0105] It can be understood that in the embodiment of the present invention, the first reference obstacle and the second reference obstacle are introduced to determine the first reference line and the second reference line, and then the obstacle angle is determined, thereby avoiding the situation where the obstacle angle is inaccurate when the current obstacle to be detected is far away from the seed obstacle, and the accuracy of the determined obstacle angle is improved; at the same time, when it is determined according to the angle detection strategy that the current obstacle to be detected is not the target obstacle, the line fitting detection strategy is introduced to perform a second detection on the current obstacle to be detected, thereby improving the accuracy of fence detection.

[0106] In an optional embodiment, according to the line fitting detection strategy, fence detection is performed on the current obstacle to be detected to determine whether the current obstacle to be detected is a target obstacle, including: performing straight line fitting on the obstacle center points of each fence obstacle in the fence obstacle list to obtain an obstacle fitting line; determining a fitting distance between the current obstacle to be detected and the obstacle fitting line, and determining whether the current obstacle to be detected is a target obstacle based on the fitting distance and a preset fitting distance threshold.

[0107] The obstacle fitting line refers to a straight line obtained by performing straight line fitting on the obstacle center point of the fence obstacle. The embodiment of the present invention does not impose any limitation on the straight line fitting method, which can be set by a technician based on experience or needs. Exemplarily, the straight line fitting method can be a random sampling consistency method with noise resistance.

[0108] The fitting distance refers to the straight-line distance between the obstacle to be detected and the obstacle fitting line. For example, the distance between the obstacle center point of the obstacle to be detected and the obstacle fitting line is used as the fitting distance. The embodiment of the present invention does not impose any limitation on the size of the preset fitting distance threshold, which can be set by technicians according to experience or needs.

[0109] Exemplarily, judging whether the obstacle to be detected is a target obstacle based on the fitting distance and a preset fitting distance threshold includes: if the fitting distance is less than the preset fitting distance threshold, determining that the obstacle to be detected is a target obstacle; if the fitting distance is greater than or equal to the preset fitting distance threshold, prohibiting the obstacle to be detected from being a target obstacle.

[0110] It can be understood that by introducing the obstacle fitting line, determining the fitting distance between the current obstacle to be detected and the obstacle fitting line, and determining whether the current obstacle to be detected is the target obstacle based on a preset fitting distance threshold, the accuracy of determining the target obstacle is improved.

[0111] For example, if there are more than two fence obstacles in the fence obstacle list, the calculated center arrangement of the fence obstacles may be misaligned due to the centimeter-level accuracy of the laser radar. Therefore, judging only by the obstacle angle mentioned above may result in a mismatch due to the left and right misalignment of the first two fence obstacles. Therefore, it is necessary to introduce a line fitting detection strategy to allow for a certain error. The specific method is: first, use the obstacle angle to make a judgment. If the obstacle angle is less than the preset angle threshold, put it in the fence obstacle list and continue to traverse the next obstacle to be detected; if the obstacle angle is greater than or equal to the preset angle threshold, enter the line fitting judgment. The line fitting judgment is to perform straight line fitting according to the obstacle center point (x, y) value of the fence obstacle in the fence obstacle list to obtain the obstacle fitting line; calculate the fitting distance between the obstacle center point of the current obstacle to be detected and the obstacle fitting line. If the fitting distance is less than the preset fitting distance threshold, the current obstacle to be detected is added to the fence obstacle list; if not, the current obstacle to be detected is not the target obstacle, jump out of the current loop, and traverse the next obstacle to be detected.

[0112] In the embodiment of the present invention, reverse traversal can also be performed, because the fence exists not only in front of the current vehicle, but also in the rear, so it is necessary to continue to traverse backward. Since the front has been sorted, it is only necessary to traverse the obstacles to be detected that are ranked earlier than the seed obstacles. The specific traversal process is consistent with the forward traversal and will not be repeated.

[0113] S270: Perform barrier detection on the current obstacle to be detected according to the target barrier detection strategy, and determine whether the current obstacle to be detected is a target obstacle. If so, add the target obstacle to the barrier obstacle list.

[0114] S280: After all obstacles to be detected have completed barrier detection, determine barriers on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list.

[0115] Exemplarily, the barrier obstacles in the barrier obstacle list are merged into one obstacle, which is the final barrier obstacle, and other operations such as subsequent filtering can be performed according to requirements.

[0116] In the embodiment of the present invention, if the number of determined seed obstacles is at least two, a fence obstacle list corresponding to each sub-obstacle is constructed respectively, and the road fence obstacle list corresponding to each sub-obstacle is filled using the road fence detection method provided in the above embodiment. According to the filled fence obstacle list, the fences corresponding to each sub-obstacle are determined.

[0117] The embodiment of the present invention provides a road fence detection solution, which selects a target fence detection strategy from preset candidate fence detection strategies according to the number of obstacles, and refines it into: if the number of obstacles is 1, the target fence detection strategy is a straight line detection strategy; if the number of obstacles is 2, the target fence detection strategy is an angle detection strategy; if the number of obstacles exceeds 2, the target fence detection strategy includes an angle detection strategy and / or a line fitting detection strategy, thereby improving the determination mechanism of the target fence detection strategy. The above scheme improves the accuracy of the determined target fence detection strategy by determining the corresponding target fence detection strategy according to the number of obstacles, thereby improving the accuracy of subsequent fence detection based on the target fence detection strategy; at the same time, the embodiment of the present invention introduces a straight line detection strategy, an angle detection strategy and a line fitting detection strategy, thereby realizing fence detection of the current obstacle to be detected from multiple dimensions, and improving the diversity of candidate fence detection strategies.

[0118] On the basis of the above technical solutions, the embodiment of the present invention proposes a fence filtering algorithm based on the LiDAR point cloud segmentation algorithm, which can also be understood as a fence detection algorithm. The present invention can alleviate the drawbacks of the traditional segmentation algorithm that relies too much on the map for fence filtering, and can also ensure that the fence does not directly output and cause false detection when the fence exists in the map, causing the autonomous driving vehicle to brake or turn urgently, creating a safety hazard. The present invention can be directly added as a submodule in the target filtering stage, which is plug-and-play and does not require excessive resource consumption.

[0119] The main idea of ​​the embodiment of the present invention is to first select a long and narrow target that is relatively close to the current vehicle as a seed obstacle, and then sequentially merge distant discrete obstacles to be detected into the seed obstacle, thereby completing the segmentation and filtering of the entire fence.

[0120] Intelligent driving system includes perception, planning and control, positioning and Figure 3 A large module, for example, radar perception can obtain lidar point cloud data through lidar sensors, and can obtain obstacles in the current environment through positioning and map information, as well as radar point cloud processing related algorithms. For some roadside fences that exist in the map, the traditional set segmentation point cloud algorithm cannot effectively filter them and will output the fences as obstacles. In addition, since the point cloud becomes more discrete as the distance increases, a fence obstacle will be detected as multiple targets in the distance, which will have a certain impact on subsequent decision-making.

[0121] In the embodiment of the present invention, after the obstacle list perceived by the radar is output, the method proposed by the present invention is inserted to obtain the part of the obstacle list that belongs to the fence. Filtering out these fences is more conducive to path planning.

[0122] Embodiment 3

[0123] Figure 3 1 is a schematic diagram of the structure of a road barrier detection device provided in Embodiment 3 of the present invention. This embodiment is applicable to the case of detecting barriers on a road, and the method can be performed by a road barrier detection device, which can be implemented in software and / or hardware, and can be configured in an electronic device that carries a road barrier detection function.

[0124] like Figure 3 As shown, the device includes: an obstacle determination module 310, an obstacle quantity determination module 320, a detection strategy determination module 330, a fence detection module 340 and a fence determination module 350. Among them,

[0125] The obstacle determination module 310 is used to obtain an obstacle list of the current vehicle and determine a seed obstacle and an obstacle to be detected from the obstacle list;

[0126] The obstacle quantity determination module 320 is used to add the seed obstacle to the barrier obstacle list, and determine the obstacle quantity of the barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include seed obstacles and / or target obstacles;

[0127] A detection strategy determination module 330, for determining a current obstacle to be detected from the obstacles to be detected, and selecting a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles;

[0128] A barrier detection module 340 is used to perform barrier detection on the current obstacle to be detected according to the target barrier detection strategy, determine whether the current obstacle to be detected is the target obstacle, and if so, add the target obstacle to the barrier obstacle list;

[0129] The barrier determination module 350 is used to determine the barriers on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list after the barrier detection is completed for all the obstacles to be detected.

[0130] The embodiment of the present invention provides a road barrier detection scheme, which obtains the obstacle list of the current vehicle, determines the seed obstacle and the obstacle to be detected from the obstacle list; adds the seed obstacle to the barrier obstacle list, and determines the obstacle quantity of the barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include the seed obstacle and / or the target obstacle; determines the current obstacle to be detected from the obstacles to be detected, and selects the target barrier detection strategy from the preset candidate barrier detection strategies according to the obstacle quantity; performs barrier detection on the current obstacle to be detected according to the target barrier detection strategy, determines whether the current obstacle to be detected is the target obstacle, and if so, adds the target obstacle to the barrier obstacle list; after the barrier detection of all obstacles to be detected is completed, determines the barrier on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list. The above scheme improves the accuracy of barrier detection by determining the barrier obstacle according to the seed obstacle and the obstacle to be detected, and then determining the barrier on the road according to the barrier obstacle.

[0131] Optionally, if the candidate fence detection strategies include a straight line detection strategy, an angle detection strategy, and a line fitting detection strategy, the detection strategy determination module 330 includes:

[0132] A first strategy determination unit, configured to determine that if the number of obstacles is 1, the target fence detection strategy is a straight line detection strategy;

[0133] A second strategy determination unit, configured to determine, if the number of obstacles is 2, that the target fence detection strategy is an angle detection strategy;

[0134] The third strategy determination unit is used for, if the number of obstacles exceeds 2, then the target fence detection strategy includes an angle detection strategy and / or a line fitting detection strategy.

[0135] Optionally, if the target fence detection strategy is a straight line detection strategy, the fence detection module 340 includes:

[0136] A coordinate set determining unit, used to obtain a seed center point of a seed obstacle, and divide the seed coordinate points in the seed obstacle into a first coordinate set and a second coordinate set according to the seed center point;

[0137] a seed line determination unit, configured to determine a first coordinate mean and a second coordinate mean in the first coordinate set and the second coordinate set, respectively, and determine a seed obstacle line according to the first coordinate mean and the second coordinate mean;

[0138] The target obstacle judgment unit is used to determine the seed distance between the current obstacle to be detected and the seed obstacle line, and judge whether the current obstacle to be detected is the target obstacle according to the seed distance and a preset seed distance threshold.

[0139] Optionally, if the target fence detection strategy is an angle detection strategy, the fence detection module 340 includes:

[0140] an obstacle line determining unit, configured to determine a first obstacle line between the seed obstacle and the current obstacle to be detected, and to determine a second obstacle line between the seed obstacle and the target obstacle;

[0141] The target obstacle determination unit is used to determine the obstacle angle between the first obstacle line and the second obstacle line, and determine whether the current obstacle to be detected is the target obstacle according to the obstacle angle and a preset angle threshold.

[0142] Optionally, if the target fence detection strategy includes an angle detection strategy and / or a line fitting detection strategy, the fence detection module 340 includes:

[0143] a reference obstacle determination unit, configured to determine a center distance between an obstacle center point of each of the barrier obstacles in the barrier obstacle list and an obstacle center point of the obstacle to be detected currently, and select a first reference obstacle and a second reference obstacle from the barrier obstacles according to the center distance;

[0144] a reference line determining unit, configured to determine a first reference line between the first reference obstacle and the obstacle currently to be detected, and to determine a second reference line between the first reference obstacle and the second reference obstacle;

[0145] an obstacle detection unit, configured to determine an obstacle angle between the first reference line and the second reference line, and determine whether the obstacle to be detected is the target obstacle according to the obstacle angle and a preset angle threshold;

[0146] The barrier detection unit is used to, if not, perform barrier detection on the obstacle to be detected according to the line fitting detection strategy to determine whether the obstacle to be detected is the target obstacle.

[0147] Optionally, the fence detection unit is specifically used to:

[0148] Performing straight line fitting on the obstacle center point of each of the barrier obstacles in the barrier obstacle list to obtain an obstacle fitting line;

[0149] A fitting distance between the current obstacle to be detected and the obstacle fitting line is determined, and according to the fitting distance and a preset fitting distance threshold, whether the current obstacle to be detected is the target obstacle is determined.

[0150] Optionally, the obstacle determination module 310 includes:

[0151] an obstacle to be screened determining unit, configured to determine an obstacle to be screened from the candidate obstacles according to obstacle attribute parameters of each candidate obstacle in the obstacle list and a preset obstacle screening strategy;

[0152] A sorting result determining unit, used for sorting the obstacles to be screened according to the obstacle center points of the obstacles to be screened, to obtain an obstacle sorting result;

[0153] A seed obstacle determination unit, configured to determine a seed obstacle from the obstacles to be screened according to obstacle attribute parameters of each obstacle to be screened and a preset seed determination strategy;

[0154] The obstacle to be detected determining unit is used to determine the obstacle to be detected according to the seed obstacle and the obstacle sorting result.

[0155] The road barrier detection device provided in the embodiment of the present invention can execute the road barrier detection method provided in any embodiment of the present invention, and has the functional modules and beneficial effects corresponding to executing each road barrier detection method.

[0156] In the technical solution of the present invention, the collection, storage, use, processing, transmission, provision and disclosure of the obstacle list, candidate fence detection strategy, obstacle screening strategy and seed determination strategy involved all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0157] Embodiment 4

[0158] Figure 4 4 is a schematic diagram of the structure of an electronic device for implementing a road barrier detection method provided by Embodiment 4 of the present invention. The electronic device 410 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0159] like Figure 4 As shown, the electronic device 410 includes at least one processor 411, and a memory connected to the at least one processor 411 in communication, such as a read-only memory (ROM) 412, a random access memory (RAM) 413, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 411 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 412 or the computer program loaded from the storage unit 418 to the random access memory (RAM) 413. In the RAM 413, various programs and data required for the operation of the electronic device 410 can also be stored. The processor 411, the ROM 412, and the RAM 413 are connected to each other via a bus 414. An input / output (I / O) interface 415 is also connected to the bus 414.

[0160] Multiple components in the electronic device 410 are connected to the I / O interface 415, including: an input unit 416, such as a keyboard, a mouse, etc.; an output unit 417, such as various types of displays, speakers, etc.; a storage unit 418, such as a disk, an optical disk, etc.; and a communication unit 419, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 419 allows the electronic device 410 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0161] The processor 411 may be a variety of general and / or dedicated processing components with processing and computing capabilities. Some examples of the processor 411 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 411 executes the various methods and processes described above, such as a road barrier detection method.

[0162] In some embodiments, the road barrier detection method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 418. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 410 via the ROM 412 and / or the communication unit 419. When the computer program is loaded into the RAM 413 and executed by the processor 411, one or more steps of the road barrier detection method described above may be performed. Alternatively, in other embodiments, the processor 411 may be configured to perform the road barrier detection method in any other appropriate manner (e.g., by means of firmware).

[0163] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0164] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0165] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer 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 foregoing.

[0166] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0167] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0168] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0169] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0170] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A road barrier detection method, It is characterized in that include: Obtaining an obstacle list of the current vehicle, and determining a seed obstacle and an obstacle to be detected from the obstacle list; Adding the seed obstacle to a barrier obstacle list, and determining the number of barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include seed obstacles and / or target obstacles; Determine the current obstacle to be detected from the obstacles to be detected, and select a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles; According to the target barrier detection strategy, barrier detection is performed on the current obstacle to be detected to determine whether the current obstacle to be detected is the target obstacle, and if so, the target obstacle is added to the barrier obstacle list; After the barriers of the obstacles to be detected are all detected, the barriers on the road where the current vehicle is located are determined according to the barrier obstacles in the barrier obstacle list.

2. The method according to claim 1, It is characterized in that If the candidate barrier detection strategies include a straight line detection strategy, an angle detection strategy, and a line fitting detection strategy, then selecting a target barrier detection strategy from the preset candidate barrier detection strategies according to the number of obstacles includes: If the number of obstacles is 1, the target fence detection strategy is a straight line detection strategy; If the number of obstacles is 2, the target fence detection strategy is the angle detection strategy; If the number of obstacles exceeds 2, the target fence detection strategy includes an angle detection strategy and / or a line fitting detection strategy.

3. The method according to claim 2, It is characterized in that If the target fence detection strategy is a straight line detection strategy, then performing fence detection on the current obstacle to be detected according to the target fence detection strategy to determine whether the current obstacle to be detected is the target obstacle includes: Acquire a seed center point of a seed obstacle, and divide the seed coordinate points in the seed obstacle into a first coordinate set and a second coordinate set according to the seed center point; Determine a first coordinate mean and a second coordinate mean in the first coordinate set and the second coordinate set respectively, and determine a seed obstacle line according to the first coordinate mean and the second coordinate mean; A seed distance between the current obstacle to be detected and the seed obstacle line is determined, and based on the seed distance and a preset seed distance threshold, it is determined whether the current obstacle to be detected is the target obstacle.

4. The method according to claim 2, It is characterized in that If the target barrier detection strategy is an angle detection strategy, then performing barrier detection on the current obstacle to be detected according to the target barrier detection strategy to determine whether the current obstacle to be detected is the target obstacle includes: Determine a first obstacle line between the seed obstacle and the current obstacle to be detected, and determine a second obstacle line between the seed obstacle and the target obstacle; An obstacle angle between the first obstacle line and the second obstacle line is determined, and based on the obstacle angle and a preset angle threshold, it is determined whether the obstacle to be detected is the target obstacle.

5. The method according to claim 2, It is characterized in that If the target barrier detection strategy includes an angle detection strategy and / or a line fitting detection strategy, then performing barrier detection on the current obstacle to be detected according to the target barrier detection strategy to determine whether the current obstacle to be detected is the target obstacle includes: Determine a center distance between an obstacle center point of each of the barrier obstacles in the barrier obstacle list and an obstacle center point of the obstacle to be detected currently, and select a first reference obstacle and a second reference obstacle from the barrier obstacles according to the center distance; Determining a first reference line between the first reference obstacle and the obstacle currently to be detected, and determining a second reference line between the first reference obstacle and the second reference obstacle; Determine an obstacle angle between the first reference line and the second reference line, and determine whether the obstacle to be detected is the target obstacle according to the obstacle angle and a preset angle threshold; If not, fence detection is performed on the current obstacle to be detected according to the line fitting detection strategy to determine whether the current obstacle to be detected is the target obstacle.

6. The method according to claim 5, It is characterized in that The step of performing fence detection on the obstacle to be detected according to the line fitting detection strategy to determine whether the obstacle to be detected is the target obstacle includes: Performing straight line fitting on the obstacle center point of each of the barrier obstacles in the barrier obstacle list to obtain an obstacle fitting line; A fitting distance between the current obstacle to be detected and the obstacle fitting line is determined, and according to the fitting distance and a preset fitting distance threshold, whether the current obstacle to be detected is the target obstacle is determined.

7. The method according to any one of claims 1 to 6, It is characterized in that The step of determining the seed obstacle and the obstacle to be detected from the obstacle list includes: Determining obstacles to be screened from the candidate obstacles according to obstacle attribute parameters of each candidate obstacle in the obstacle list and a preset obstacle screening strategy; Sorting the obstacles to be screened according to the obstacle center points of the obstacles to be screened to obtain an obstacle sorting result; Determining seed obstacles from the obstacles to be screened according to obstacle attribute parameters of each obstacle to be screened and a preset seed determination strategy; The obstacle to be detected is determined according to the seed obstacle and the obstacle sorting result.

8. A road barrier detection device, It is characterized in that include: An obstacle determination module is used to obtain an obstacle list of the current vehicle and determine a seed obstacle and an obstacle to be detected from the obstacle list; An obstacle quantity determination module, used for adding the seed obstacle to the barrier obstacle list, and determining the obstacle quantity of the barrier obstacles in the barrier obstacle list; wherein the barrier obstacles include seed obstacles and / or target obstacles; A detection strategy determination module, used to determine the current obstacle to be detected from the obstacles to be detected, and select a target barrier detection strategy from preset candidate barrier detection strategies according to the number of obstacles; A barrier detection module, configured to perform barrier detection on the obstacle to be detected according to the target barrier detection strategy, determine whether the obstacle to be detected is the target obstacle, and if so, add the target obstacle to the barrier obstacle list; The barrier determination module is used to determine the barriers on the road where the current vehicle is located according to the barrier obstacles in the barrier obstacle list after the barrier detection of the obstacles to be detected is completed.

9. An electronic device, It is characterized in that include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement a road barrier detection method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, a road barrier detection method as described in any one of claims 1 to 7 is implemented.

Images