An obstacle detection method, apparatus, device and medium

Abstract

The application discloses an obstacle detection method, device and equipment and a storage medium, and comprises the following steps: acquiring a target frame point cloud obtained by a millimeter wave radar when detecting a target obstacle, and determining a target point cloud cluster of the target frame point cloud and a previous point cloud cluster of a previous frame point cloud of the target frame point cloud by using a clustering algorithm; determining a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the previous point cloud cluster; mapping the previous point cloud cluster to the target point cloud cluster according to the mapping relationship, and determining an updated radial velocity corresponding to the mapped previous point cloud cluster; determining an aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster by using the clustering algorithm according to the updated radial velocity and the target radial velocity, and determining obstacle information of the target obstacle according to the aggregated point cloud cluster. The aggregated point cloud cluster with high point cloud density can be obtained, and the accuracy of the obstacle information is improved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of autonomous driving vehicles, and in particular to a method and device for obstacle detection, an apparatus and a storage medium. BACKGROUND

[0002] With the development of autonomous driving vehicles, 4D millimeter wave radars are currently used to detect obstacles during the driving process of vehicles. At present, when obstacles are detected by 4D millimeter wave radars, the number of effective point clouds output is small, and the density of point clouds is low. Detecting obstacles through point cloud clustering results will lead to unstable detection results of the position information and size information of obstacles, and will also cause the problem of point cloud over-segmentation. The current obstacle detection technology cannot meet the needs of L3 and above autonomous driving vehicles for obstacle detection. Therefore, how to improve the accuracy of the detection results of the position information and size information of obstacles around a vehicle by 4D millimeter wave radars is a problem to be solved. SUMMARY

[0003] The present application provides a method and device for obstacle detection, an apparatus and a storage medium, which can improve the accuracy of the detection results of the position information and size information of obstacles around a vehicle by 4D millimeter wave radars.

[0004] According to an aspect of the present application, a method for obstacle detection is provided, comprising:

[0005] acquiring a target frame point cloud obtained when a millimeter wave radar detects a target obstacle, and using a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and an upper point cloud cluster of a previous frame point cloud of the target frame point cloud;

[0006] determining a mapping relationship between the target point cloud cluster and the upper point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and an upper radial velocity corresponding to the upper point cloud cluster;

[0007] mapping the upper point cloud cluster to the target point cloud cluster according to the mapping relationship, and determining an updated radial velocity corresponding to the mapped upper point cloud cluster;

[0008] determining an aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster by the clustering algorithm according to the updated radial velocity and the target radial velocity, and determining obstacle information of the target obstacle according to the aggregated point cloud cluster.

[0009] According to another aspect of the present application, a device for obstacle detection is provided, comprising:

[0010] The point cloud cluster determination module is configured to acquire a target frame point cloud acquired when the millimeter wave radar detects a target obstacle, and determine a target point cloud cluster of the target frame point cloud and a previous point cloud cluster of a previous frame point cloud of the target frame point cloud by using a clustering algorithm.

[0011] The mapping relationship determination module is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the previous point cloud cluster.

[0012] The point cloud cluster mapping module is configured to map the previous point cloud cluster to the target point cloud cluster according to the mapping relationship, and determine an updated radial velocity corresponding to the mapped previous point cloud cluster.

[0013] The obstacle information determination module is configured to determine an aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster by using the clustering algorithm according to the updated radial velocity and the target radial velocity, and determine obstacle information of the target obstacle according to the aggregated point cloud cluster.

[0014] According to another aspect of the present application, an electronic device is provided, which comprises:

[0015] at least one processor; and

[0016] a memory connected to the at least one processor in communication; wherein

[0017] the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the obstacle detection method according to any one of the embodiments of the present application.

[0018] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to execute the obstacle detection method according to any one of the embodiments of the present application when executed by the processor.

[0019] The technical scheme of the embodiment of the present application obtains a target frame point cloud obtained by a millimeter wave radar when detecting a target obstacle, adopts a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and a previous point cloud cluster of a previous frame point cloud of the target frame point cloud, determines a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the previous point cloud cluster, maps the previous point cloud cluster to the target point cloud cluster according to the mapping relationship and determines an updated radial velocity corresponding to the mapped previous point cloud cluster, and determines an aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster by the clustering algorithm according to the updated radial velocity and the target radial velocity, and determines obstacle information of the target obstacle according to the aggregated point cloud cluster. The above scheme solves the problem that when the obstacle is detected by the 4D millimeter wave radar, the effective point cloud quantity output is small and the point cloud density is low, the obstacle information of the target obstacle is determined according to the point cloud clustering result, and the detection result of the obstacle information is unstable. The above scheme determines the point cloud clusters of multiple frames of point clouds respectively by the aggregation algorithm, aggregates the multiple frames of point cloud clusters according to the mapping relationship between the point cloud clusters of the multiple frames of point clouds, determines the aggregated point cloud cluster, and determines the obstacle information of the target obstacle according to the aggregated point cloud cluster. In the case that the hardware of the single 4D millimeter wave radar does not change, the point cloud density and quality are improved by the multi-frame fusion algorithm, in the case of the same hardware cost, the aggregated point cloud cluster with high point cloud density can be obtained, the obstacle information of the target obstacle is determined according to the aggregated point cloud cluster, and the accuracy of the obstacle information can be improved.

[0020] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0022] Figure 1 A flowchart of a method for detecting an obstacle provided for the first embodiment of the present application;

[0023] Figure 2 A flowchart of a method for detecting an obstacle provided for the second embodiment of the present application;

[0024] Figure 3 A flowchart of a method for detecting an obstacle provided for the third embodiment of the present application;

[0025] Figure 4 A structural schematic diagram of an obstacle detection device provided for Embodiment Four of the present application;

[0026] Figure 5 A structural schematic diagram of an electronic device provided for Embodiment Five of the present application. DETAILED DESCRIPTION

[0027] In order to enable persons skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative labor should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "candidate" and "target" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "etc." and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0029] Embodiment One

[0030] Figure 1 A flowchart of an obstacle detection method provided for Embodiment One of the present application is provided, and the present embodiment can be applicable to detecting target obstacles around an autonomous vehicle to obtain obstacle information. The method can be performed by an obstacle detection device, which can be realized in the form of hardware and / or software, and the obstacle detection device can be configured in an electronic device. As shown in the figure, the method comprises: Figure 1

[0031] S110, obtaining a target frame point cloud obtained when a millimeter wave radar detects a target obstacle, and using a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and an upper point cloud cluster of a previous frame point cloud of the target frame point cloud.

[0032] ​The millimeter wave radar refers to a 4D millimeter wave radar installed on an automatic driving vehicle, and the 4D millimeter wave radar can analyze distance, azimuth, speed and height dimension data of a target obstacle, and can perceive information of the target obstacle in four dimensions. The target frame point cloud refers to point cloud data obtained by scanning the surrounding environment by the millimeter wave radar. One frame of point cloud refers to a point cloud set formed by one scanning cycle of the millimeter wave radar. The point cloud cluster of the point cloud data can be obtained by clustering the point cloud data. The clustering algorithm can be a Dbscan clustering algorithm.

[0033] Specifically, network data of the millimeter wave radar in detecting the target obstacle is obtained, the network data of the millimeter wave radar is parsed into a single frame of point cloud according to a data protocol for storage, and a target frame of point cloud is read from the parsed single frame of point cloud. The target frame of point cloud is analyzed by using the Dbscan clustering algorithm to determine a target point cloud cluster of the target frame of point cloud. Meanwhile, a previous frame of point cloud of the target frame of point cloud is determined, and the previous frame of point cloud is analyzed by using the Dbscan clustering algorithm to determine a previous point cloud cluster of the previous frame of point cloud.

[0034] Exemplarily, the target point cloud cluster of the target frame of point cloud and the previous point cloud cluster of the previous frame of point cloud of the target frame of point cloud can be determined through the following sub-steps:

[0035] S1101, determining a target point cloud distance function according to a target point cloud coordinate of the target frame of point cloud, a target point cloud azimuth angle difference value, and a point cloud radial velocity of the target frame of point cloud.

[0036] Exemplarily, a calculation formula of the target point cloud distance function between two point clouds in the target frame of point cloud is shown in formula (1):

[0037]

[0038] wherein dis(p1, p2) is the target point cloud distance function; w1 is a distance weight; w2 is a speed weight; (x1, y1, z1) and (x2, y2, z2) are physical coordinates of two point clouds in the target frame of point cloud; v1 and v2 are corresponding radial velocities of the two point clouds in the target frame of point cloud; Δθ is an azimuth angle difference value between the two point clouds in the target frame of point cloud; and a is a normalization coefficient.

[0039] S1102, determining a previous point cloud distance function according to a previous point cloud coordinate of the target point cloud cluster, a previous point cloud azimuth angle difference value, and a point cloud radial velocity of the previous frame of point cloud.

[0040] Exemplarily, a calculation formula of the previous point cloud distance function between two point clouds in the previous frame of point cloud is shown in formula (2):

[0041]

[0042] wherein dis(p1', p2') is the last point cloud distance function; (x1', y1', z1') and (x2', y2', z2') are physical coordinates of two point clouds in the last frame point cloud; v1' and v2' are radial velocities corresponding to the two point clouds in the last frame point cloud; and Δθ' is the azimuth angle difference between the two point clouds in the last frame point cloud.

[0043] S1103, determining the target point cloud cluster of the target frame point cloud and the last point cloud cluster of the last frame point cloud based on the target point cloud distance function and the last point cloud distance function through a clustering algorithm.

[0044] Specifically, each target point cloud cluster of the target frame point cloud is determined based on the target point cloud distance function through a Dbscan clustering algorithm. Each last point cloud cluster of the last frame point cloud is determined based on the last point cloud distance function through the Dbscan clustering algorithm.

[0045] It can be understood that the above scheme provides a scheme for determining a target point cloud distance function of a target point cloud cluster and a last point cloud distance function of a last point cloud cluster, which can ensure that each point cloud in the determined target point cloud cluster and last point cloud cluster avoids the influence of micro-Doppler, thereby avoiding errors in radial velocity calculation and improving the calculation accuracy of the target radial velocity corresponding to the target point cloud cluster and the last radial velocity corresponding to the last point cloud cluster.

[0046] S120, determining a mapping relationship between the target point cloud cluster and the last point cloud cluster according to the target radial velocity corresponding to the target point cloud cluster and the last radial velocity corresponding to the last point cloud cluster.

[0047] wherein the target radial velocity corresponding to each target point cloud cluster is the average of the radial velocities corresponding to the point clouds in each target point cloud cluster. The last radial velocity corresponding to each last point cloud cluster is the average of the radial velocities corresponding to the point clouds in each last point cloud cluster.

[0048] Specifically, the target point cloud cluster and the last point cloud cluster are matched according to the target radial velocity corresponding to each target point cloud cluster and the last radial velocity corresponding to each last point cloud cluster, to determine the target point cloud cluster having a mapping relationship with the last point cloud cluster.

[0049] S130, mapping the last point cloud cluster to the target point cloud cluster according to the mapping relationship, and determining an updated radial velocity corresponding to the mapped last point cloud cluster.

[0050] wherein the updated radial velocity refers to the radial velocity corresponding to the mapped last point cloud cluster after the last point cloud cluster is mapped to the target point cloud cluster.

[0051] Specifically, according to the mapping relationship between the target point cloud cluster and the upper point cloud cluster, each upper point cloud cluster is mapped into the target point cloud cluster having a mapping relationship with the upper point cloud cluster, and after the mapping is completed, the updated radial velocity corresponding to the upper point cloud cluster mapped into the target point cloud cluster is determined.

[0052] In S140, according to the updated radial velocity and the target radial velocity, the aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster is determined through a clustering algorithm, and the obstacle information of the target obstacle is determined according to the aggregated point cloud cluster.

[0053] The aggregated point cloud cluster refers to the point cloud cluster obtained by clustering analysis on the target point cloud cluster and the upper point cloud cluster mapped into the target point cloud cluster. The obstacle information includes the obstacle size and the obstacle contour.

[0054] Specifically, according to the updated radial velocity and the target radial velocity, the aggregated point cloud distance function of the target point cloud cluster and the mapped upper point cloud cluster is determined. Based on the aggregated point cloud distance function, the target point cloud cluster and the upper point cloud cluster mapped into the target point cloud cluster are analyzed by a clustering algorithm to determine the aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster. The bounding box is determined according to the aggregated point cloud cluster, and the obstacle size and the obstacle contour are determined according to the point cloud in the bounding box, so that the autonomous vehicle can plan an obstacle avoidance driving route based on the obstacle size and the obstacle contour.

[0055] For example, the method for determining the aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster can be: according to the target point cloud coordinates of the target frame point cloud, the target point cloud azimuth angle difference value, the updated radial velocity, the target radial velocity, the mapped point cloud coordinates of the mapped upper point cloud cluster, and the mapped point cloud azimuth angle difference value, the aggregated point cloud distance function of the target point cloud cluster and the mapped upper point cloud cluster is determined; based on the aggregated point cloud distance function, the aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster is determined through a clustering algorithm.

[0056] The above scheme determines the aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster through the aggregated point cloud distance function and the clustering algorithm, which can improve the stability of the aggregated point cloud cluster.

[0057] The technical scheme provided by the embodiment acquires target frame point clouds obtained when a millimeter wave radar detects a target obstacle, adopts a clustering algorithm to determine a target point cloud cluster of the target frame point clouds and an upper point cloud cluster of a previous frame point cloud of the target frame point clouds, determines a mapping relationship between the target point cloud cluster and the upper point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the upper point cloud cluster, maps the upper point cloud cluster to the target point cloud cluster according to the mapping relationship and determines an updated radial velocity corresponding to the mapped upper point cloud cluster, determines an aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster through the clustering algorithm according to the updated radial velocity and the target radial velocity, and determines obstacle information of the target obstacle according to the aggregated point cloud cluster. The above scheme solves the problem that when an obstacle is detected through a 4D millimeter wave radar, the number of effective point clouds output is small and the point cloud density is low, the obstacle information of the target obstacle is determined through the point cloud clustering result, and the detection result of the obstacle information is unstable. The above scheme determines point cloud clusters of multiple frames of point clouds through an aggregation algorithm, aggregates the multiple frames of point cloud clusters according to the mapping relationship between the point cloud clusters of the multiple frames of point clouds, determines an aggregated point cloud cluster, and determines the obstacle information of the target obstacle according to the aggregated point cloud cluster. In the case that the hardware of a single 4D millimeter wave radar does not change, the point cloud density and quality are improved through a multi-frame fusion algorithm, in the case of the same hardware cost, the aggregated point cloud cluster with high point cloud density can be obtained, the obstacle information of the target obstacle is determined through the aggregated point cloud cluster, and the accuracy of the obstacle information can be improved.

[0058] Embodiment Two

[0059] Figure 2 The flowchart of the obstacle detection method provided for the second embodiment of the application is optimized on the basis of the above-mentioned embodiments, and a preferred implementation manner of determining the mapping relationship between the target point cloud cluster and the upper point cloud cluster according to the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the upper point cloud cluster is given. Specifically, as shown in Figure 2 The method comprises the following steps.

[0060] In S210, target frame point clouds obtained when a millimeter wave radar detects a target obstacle are acquired, and a clustering algorithm is adopted to determine a target point cloud cluster of the target frame point clouds and an upper point cloud cluster of a previous frame point cloud of the target frame point clouds.

[0061] In S220, a point cloud mapping distance between the target point cloud cluster and the upper point cloud cluster is determined according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the upper point cloud cluster.

[0062] Specifically, the average point cloud coordinate of the target point cloud cluster is taken as the target point cloud cluster coordinate, and the average point cloud coordinate of the previous point cloud cluster is taken as the previous point cloud cluster coordinate. According to the target point cloud cluster coordinate and the previous point cloud cluster coordinate, the obstacle running speed of the target obstacle is determined. According to the obstacle running speed of the target obstacle, the target radial speed corresponding to the target point cloud cluster, and the previous radial speed corresponding to the previous point cloud cluster, the point cloud mapping distance between the target point cloud cluster and the previous point cloud cluster is determined.

[0063] For example, the calculation formula of the point cloud mapping distance between the target point cloud cluster and the previous point cloud cluster is shown in formula (3):

[0064] dis(A1, B1) = λ1·|v′ a -v|+ λ2·|v′ b -v| (3)

[0065] dis(A1, B1) is the point cloud mapping distance between the target point cloud cluster and the previous point cloud cluster; v′ a is the target radial speed corresponding to the target point cloud cluster; v′ b is the previous radial speed corresponding to the previous point cloud cluster; λ1 represents the speed error weight of the target point cloud cluster; λ1 represents the speed error weight of the previous point cloud cluster; v represents the obstacle running speed.

[0066] S230, determining a mapping distance matrix according to the point cloud mapping distance, and determining a target mapping distance of each row in the mapping distance matrix.

[0067] The mapping distance matrix is a matrix generated according to the point cloud mapping distance, and each matrix element in the mapping distance matrix is a point cloud mapping distance.

[0068] Specifically, after obtaining the point cloud mapping distance between all target point cloud clusters and previous point cloud clusters, a mapping distance matrix is generated according to the obtained point cloud mapping distance. The number of rows of the mapping distance matrix is consistent with the number of previous point cloud clusters, and the number of columns of the mapping distance matrix is consistent with the number of target point cloud clusters. The same row of matrix data in the mapping distance matrix is the point cloud mapping distance of a previous point cloud cluster mapped to each target point cloud cluster. The smallest point cloud mapping distance in each row of the mapping distance matrix is determined as the target mapping distance corresponding to the row.

[0069] S240, determining a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target mapping distance.

[0070] Specifically, according to the preset mapping distance threshold, the target mapping distance is screened, and the target mapping distance less than the mapping distance threshold is determined as an effective mapping distance. If the target mapping distance between the target point cloud cluster and the previous point cloud cluster is the effective mapping distance, it is determined that there is a mapping relationship between the target point cloud cluster and the previous point cloud cluster.

[0071] For example, the method for determining the mapping relationship between the target point cloud cluster and the previous point cloud cluster can be: determining the mapping distance threshold according to the inter-frame timestamp, the speed measurement accuracy of the millimeter wave radar, and the ranging accuracy; screening the target mapping distance based on the mapping distance threshold to determine the effective mapping distance; and determining the mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the effective mapping distance.

[0072] Specifically, the calculation formula of the mapping distance threshold is shown in formula (4):

[0073] e max = (e speed + e dis / Δt)*4 (4)

[0074] Wherein, e max is the mapping distance threshold; e speed is the speed measurement accuracy of the millimeter wave radar; and Δt is the inter-frame timestamp.

[0075] The target mapping distance is compared with the mapping distance threshold, and the target mapping distance less than the mapping distance threshold is determined as the effective mapping distance. It is determined that there is a mapping relationship between the previous point cloud cluster corresponding to the effective mapping distance and the target point cloud cluster.

[0076] The above scheme provides a method for determining a mapping distance threshold. According to the mapping distance threshold, the target mapping distance is screened to obtain an effective mapping distance for determining the mapping relationship between the target point cloud cluster and the previous point cloud cluster, which can improve the accuracy of the determined mapping relationship.

[0077] S250, according to the mapping relationship, mapping the previous point cloud cluster to the target point cloud cluster, and determining the updated radial velocity corresponding to the mapped previous point cloud cluster.

[0078] S260, according to the updated radial velocity and the target radial velocity, the clustering algorithm is used to determine the aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster, and the obstacle information of the target obstacle is determined according to the aggregated point cloud cluster.

[0079] The technical scheme of the embodiment obtains a target frame point cloud obtained when a millimeter wave radar detects a target obstacle, adopts a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and an upper point cloud cluster of a previous frame point cloud of the target frame point cloud, determines a point cloud mapping distance between the target point cloud cluster and the upper point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the upper point cloud cluster, determines a mapping distance matrix according to the point cloud mapping distance and determines a target mapping distance of each row in the mapping distance matrix, determines a mapping relationship between the target point cloud cluster and the upper point cloud cluster according to the target mapping distance, maps the upper point cloud cluster to the target point cloud cluster according to the mapping relationship and determines an updated radial velocity corresponding to the mapped upper point cloud cluster, and determines an aggregated point cloud cluster of the target point cloud cluster and the mapped upper point cloud cluster by the clustering algorithm according to the updated radial velocity and the target radial velocity and determines obstacle information of the target obstacle according to the aggregated point cloud cluster. The number of target point cloud clusters and the number of upper point cloud clusters may be inconsistent, so the target point cloud cluster and the upper point cloud cluster may not have a one-to-one correspondence, and when the mapping relationship between the target point cloud cluster and the upper point cloud cluster is determined, a phenomenon that one upper point cloud cluster has a mapping relationship with multiple target point cloud clusters may occur, resulting in inaccurate mapping relationship. The above scheme forms a mapping distance matrix according to the point cloud mapping distance between the target point cloud cluster and the upper point cloud cluster, determines the mapping relationship between the target point cloud cluster and the upper point cloud cluster according to the mapping distance matrix, can ensure that each upper point cloud cluster can only have a mapping relationship with one target point cloud cluster, improves the accuracy of the mapping relationship, avoids the problem of mapping error of the upper point cloud cluster in the future, further improves the reliability of the aggregated point cloud cluster, and thus improves the accuracy of the determined obstacle information of the target obstacle.

[0080] Embodiment three

[0081] Figure 3 A flowchart of a method for detecting an obstacle provided by the third embodiment of the application, which is optimized on the basis of the above-mentioned embodiments and gives a preferred implementation manner for determining a target radial velocity corresponding to a target point cloud cluster and a previous radial velocity corresponding to an upper point cloud cluster. Specifically, as shown in the figure, the method comprises the following steps. Figure 3

[0082] S310, obtaining a target frame point cloud obtained when a millimeter wave radar detects a target obstacle, and adopting a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and an upper point cloud cluster of a previous frame point cloud of the target frame point cloud.

[0083] ​S320, set a relative coordinate system based on the vehicle of the device millimeter wave radar, take the average point cloud coordinate of the target point cloud cluster in the relative coordinate system as the target point cloud coordinate, and take the average point cloud coordinate of the upper point cloud cluster in the relative coordinate system as the upper point cloud coordinate.

[0084] Specifically, the gravity position of the vehicle of the device millimeter wave radar can be set as the origin of the relative coordinate system, and the driving direction of the vehicle can be set as the y-axis of the relative coordinate system to construct the relative coordinate system. The average point cloud coordinate of the point cloud in the target point cloud cluster in the relative coordinate system is taken as the target point cloud coordinate, and the average point cloud coordinate of the point cloud in the upper point cloud cluster in the relative coordinate system is taken as the upper point cloud coordinate.

[0085] S330, determine the obstacle running speed of the target obstacle according to the target point cloud coordinate and the upper point cloud coordinate.

[0086] The calculation formula of the obstacle running speed of the target obstacle is shown in formula (5):

[0087]

[0088] Wherein, v is the obstacle running speed of the target obstacle; x a is the horizontal coordinate of the target point cloud coordinate; y a is the vertical coordinate of the target point cloud cluster; x b is the horizontal coordinate of the upper point cloud cluster; y b is the vertical coordinate of the upper point cloud cluster.

[0089] S340, based on the relative coordinate system, determine the target radial speed corresponding to the target point cloud cluster and the upper radial speed corresponding to the upper point cloud cluster according to the obstacle running speed.

[0090] For example, based on the relative coordinate system, the method of determining the target radial speed corresponding to the target point cloud cluster and the upper radial speed corresponding to the upper point cloud cluster according to the obstacle running speed can be: determining the first connecting line between the coordinate system origin of the relative coordinate system and the target point cloud center, the second connecting line between the coordinate system origin and the upper point cloud cluster, and the third connecting line between the target point cloud center of the target point cloud cluster and the upper point cloud center of the upper point cloud cluster, and determining the first running azimuth angle between the first connecting line and the third connecting line, and the second running azimuth angle between the second connecting line and the third connecting line; according to the obstacle running speed and the first running azimuth angle, determine the target radial speed corresponding to the target point cloud cluster; according to the obstacle running speed and the second running azimuth angle, determine the upper radial speed corresponding to the upper point cloud cluster.

[0091] For example, the calculation formula of the target radial speed corresponding to the target point cloud cluster is shown in formula (6):

[0092] v′ a = v·sinθ1 (6)

[0093] Where, v′ a θ1 is the radial velocity of the target; v is the speed of the obstacle; θ1 is the first azimuth angle between the first connecting line and the third connecting line.

[0094] The radial velocity corresponding to the previous cloud cluster is shown in formula (7):

[0095] v′ b = v·sinθ2 (7)

[0096] Where, v′ b θ1 is the radial velocity; θ2 is the second azimuth angle between the second and third connecting lines.

[0097] S350. Based on the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster, determine the mapping relationship between the target point cloud cluster and the previous point cloud cluster.

[0098] S360. Based on the mapping relationship, map the previous point cloud cluster to the target point cloud cluster, and determine the updated radial velocity corresponding to the mapped previous point cloud cluster.

[0099] S370. Based on the updated radial velocity and the target radial velocity, a clustering algorithm is used to determine the aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster, and the obstacle information of the target obstacle is determined based on the aggregated point cloud cluster.

[0100] The technical scheme of the embodiment obtains a target frame point cloud obtained when a millimeter wave radar detects a target obstacle, adopts a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and a previous point cloud cluster of a previous frame point cloud of the target frame point cloud, sets a relative coordinate system based on a vehicle of the millimeter wave radar, takes average point cloud coordinates of the target point cloud cluster in the relative coordinate system as target point cloud coordinates, and takes average point cloud coordinates of the previous point cloud cluster in the relative coordinate system as previous point cloud coordinates, determines an obstacle running speed of the target obstacle according to the target point cloud coordinates and the previous point cloud coordinates, determines a target radial speed corresponding to the target point cloud cluster and a previous radial speed corresponding to the previous point cloud cluster based on the relative coordinate system and according to the obstacle running speed, determines a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target radial speed corresponding to the target point cloud cluster and the previous radial speed corresponding to the previous point cloud cluster, maps the previous point cloud cluster to the target point cloud cluster according to the mapping relationship and determines an updated radial speed corresponding to the mapped previous point cloud cluster, and determines an aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster through the clustering algorithm according to the updated radial speed and the target radial speed, and determines obstacle information of the target obstacle according to the aggregated point cloud cluster. The above scheme provides a method for determining a target radial speed corresponding to a target point cloud cluster and a previous radial speed corresponding to a previous point cloud cluster according to target point cloud coordinates of the target point cloud cluster and previous point cloud coordinates of the previous point cloud cluster, which can improve the accuracy of the target radial speed and the previous radial speed, thereby improving the detection accuracy of the obstacle information of the target obstacle.

[0101] Embodiment four

[0102] Figure 4 A structural schematic diagram of a device for detecting an obstacle is provided for the fourth embodiment of the present application. The fourth embodiment can be applied to the case of detecting a target obstacle around an autonomous vehicle to obtain obstacle information. As shown in the figure, the device for detecting an obstacle comprises a point cloud cluster determination module 410, a mapping relationship determination module 420, a point cloud cluster mapping module 430, and an obstacle information determination module 440. Figure 4

[0103] The point cloud cluster determination module 410 is configured to obtain a target frame point cloud obtained when a millimeter wave radar detects a target obstacle, and adopt a clustering algorithm to determine a target point cloud cluster of the target frame point cloud and a previous point cloud cluster of a previous frame point cloud of the target frame point cloud.

[0104] The mapping relationship determination module 420 is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to a target radial speed corresponding to the target point cloud cluster and a previous radial speed corresponding to the previous point cloud cluster.

[0105] ​The point cloud cluster mapping module 430 is configured to map the previous point cloud cluster to the target point cloud cluster according to the mapping relationship, and determine an updated radial velocity corresponding to the mapped previous point cloud cluster;

[0106] The obstacle information determination module 440 is configured to determine, by using a clustering algorithm, an aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster according to the updated radial velocity and the target radial velocity, and determine obstacle information of the target obstacle according to the aggregated point cloud cluster.

[0107] The technical scheme provided in the embodiment is configured to obtain a target frame point cloud obtained by a millimeter wave radar when detecting a target obstacle, and determine, by using a clustering algorithm, a target point cloud cluster of the target frame point cloud and a previous point cloud cluster of a previous frame point cloud of the target frame point cloud; determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to a target radial velocity corresponding to the target point cloud cluster and a previous radial velocity corresponding to the previous point cloud cluster; map the previous point cloud cluster to the target point cloud cluster according to the mapping relationship, and determine an updated radial velocity corresponding to the mapped previous point cloud cluster; determine, by using a clustering algorithm, an aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster according to the updated radial velocity and the target radial velocity, and determine obstacle information of the target obstacle according to the aggregated point cloud cluster. The above scheme solves the problem that when an obstacle is detected by using a 4D millimeter wave radar, the number of effective point clouds output is small, the density of the point clouds is low, the obstacle information of the target obstacle is determined according to the clustering result of the point clouds, and the detection result of the obstacle information is unstable. The above scheme determines point cloud clusters of multiple frames of point clouds by using an aggregation algorithm, aggregates the multiple frames of point cloud clusters according to the mapping relationship between the point cloud clusters of the multiple frames of point clouds, determines an aggregated point cloud cluster, and determines obstacle information of the target obstacle according to the aggregated point cloud cluster. In the case where the hardware of a single 4D millimeter wave radar does not change, the point cloud density and quality are improved by using a multi-frame fusion algorithm, and in the case where the hardware cost is the same, an aggregated point cloud cluster with high point cloud density can be obtained, the obstacle information of the target obstacle is determined according to the aggregated point cloud cluster, and the accuracy of the obstacle information can be improved.

[0108] For example, the mapping relationship determination module 420 further includes:

[0109] The cloud mapping distance determination unit is configured to determine a point cloud mapping distance between the target point cloud cluster and the previous point cloud cluster according to the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster;

[0110] The target mapping distance determination unit is configured to determine a mapping distance matrix according to the point cloud mapping distance, and determine a target mapping distance of each row in the mapping distance matrix;

[0111] The mapping relationship determination unit is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target mapping distance.

[0112] The mapping relationship determination unit is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target mapping distance.

[0113] The mapping relationship determination unit is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target mapping distance.

[0114] The mapping relationship determination unit is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target mapping distance.

[0115] The mapping relationship determination unit is configured to determine a mapping relationship between the target point cloud cluster and the previous point cloud cluster according to the target mapping distance.

[0116] The point cloud cluster determination module 410 is configured to:

[0117] The point cloud cluster determination module 410 is configured to:

[0118] The point cloud cluster determination module 410 is configured to:

[0119] The point cloud cluster determination module 410 is configured to:

[0120] The obstacle information determination module 440 is configured to:

[0121] The obstacle information determination module 440 is configured to:

[0122] The obstacle information determination module 440 is configured to:

[0123] The obstacle detection device further includes:

[0124] The point cloud coordinate determination module is configured to:

[0125] The running speed determining module is configured to determine a target obstacle running speed of the target obstacle according to the target point cloud coordinate and the previous point cloud coordinate.

[0126] The radial speed determining module is configured to determine a target radial speed corresponding to the target point cloud cluster and a previous radial speed corresponding to the previous point cloud cluster according to the target obstacle running speed based on the relative coordinate system.

[0127] For example, the radial speed determining module is specifically configured to:

[0128] determine a first connecting line between a coordinate system origin of the relative coordinate system and a target point cloud center of the target point cloud cluster, a second connecting line between the coordinate system origin and a previous point cloud cluster of the previous point cloud cluster, and a third connecting line between the target point cloud center of the target point cloud cluster and a previous point cloud center of the previous point cloud cluster, and determine a first running azimuth angle between the first connecting line and the third connecting line, and a second running azimuth angle between the second connecting line and the third connecting line;

[0129] determine the target radial speed corresponding to the target point cloud cluster according to the target obstacle running speed and the first running azimuth angle;

[0130] determine the previous radial speed corresponding to the previous point cloud cluster according to the target obstacle running speed and the second running azimuth angle.

[0131] The obstacle detection device provided in the embodiment can be applied to the obstacle detection method provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0132] Embodiment five

[0133] Figure 5 A structural schematic diagram of an electronic device 10 that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present application described and / or claimed in this document.

[0134] As Figure 5As shown, the electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., communicatively connected to the at least one processor 11, where the memory stores a computer program executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0135] Various components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc., an output unit 17, such as various types of displays, a speaker, etc., a storage unit 18, such as a magnetic disk, an optical disk, etc., and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0136] The processor 11 can be various general and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose 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 11 performs various methods and processes described above, such as the obstacle detection method.

[0137] In some embodiments, the obstacle detection method can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the obstacle detection method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the obstacle detection method by any other appropriate means, such as by means of firmware.

[0138] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a load programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0139] Computer programs used to implement the processes of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer program

[0140] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store computer programs for use by or in connection with an instruction execution system, apparatus, or device. Computer-readable storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0141] To provide for interaction with a user, the systems and techniques described here can 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 a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0142] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can 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.

[0143] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can 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 large management difficulty and weak business scalability in traditional physical host and VPS service.

[0144] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be executed in parallel, executed in sequence, or executed in a different order, as long as the desired results of the present disclosure are achieved, and the present disclosure is not limited herein.

[0145] The specific embodiments described above are not intended to be limiting, and persons skilled in the art will appreciate that various modifications, combinations, sub-combinations and alternatives can be made to the specific embodiments without departing from the spirit and principles of the disclosure. Accordingly, the disclosure is not limited to the specific embodiments described above, but only by the scope of the appended claims.

Claims

1. An obstacle detection method, characterized in that, include: The target frame point cloud is acquired when the millimeter-wave radar detects the target obstacle, and a clustering algorithm is used to determine the target point cloud cluster of the target frame point cloud and the previous point cloud cluster of the previous frame point cloud of the target frame point cloud. Based on the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster, determine the mapping relationship between the target point cloud cluster and the previous point cloud cluster; Based on the mapping relationship, the previous point cloud cluster is mapped to the target point cloud cluster, and the updated radial velocity corresponding to the mapped previous point cloud cluster is determined; Based on the updated radial velocity and the target radial velocity, the clustering algorithm is used to determine the aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster, and the obstacle information of the target obstacle is determined based on the aggregated point cloud cluster.

2. The method according to claim 1, characterized in that, Based on the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster, the mapping relationship between the target point cloud cluster and the previous point cloud cluster is determined, including: Based on the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster, determine the point cloud mapping distance between the target point cloud cluster and the previous point cloud cluster; The mapping distance matrix is ​​determined based on the point cloud mapping distance, and the target mapping distance in each row of the mapping distance matrix is ​​determined. Based on the target mapping distance, the mapping relationship between the target point cloud cluster and the previous point cloud cluster is determined.

3. The method according to claim 2, characterized in that, Based on the target mapping distance, the mapping relationship between the target point cloud cluster and the previous point cloud cluster is determined, including: The mapping distance threshold is determined based on the inter-frame timestamp, the velocity measurement accuracy, and the ranging accuracy of the millimeter-wave radar. The target mapping distances are filtered based on the mapping distance threshold to determine the effective mapping distances; Based on the effective mapping distance, the mapping relationship between the target point cloud cluster and the previous point cloud cluster is determined.

4. The method according to claim 1, characterized in that, Clustering algorithms are used to determine the target point cloud clusters of the target frame point cloud and the previous point cloud cluster of the previous frame point cloud, including: The target point cloud distance function is determined based on the target point cloud coordinates, the target point cloud azimuth difference, and the target frame point cloud radial velocity. The distance function of the previous point cloud is determined based on the coordinates of the previous point cloud in the target point cloud cluster, the azimuth difference of the previous point cloud, and the radial velocity of the point cloud in the previous frame. Based on the target point cloud distance function and the previous point cloud distance function, a clustering algorithm is used to determine the target point cloud cluster of the target frame point cloud and the previous point cloud cluster of the previous frame point cloud.

5. The method according to claim 1, characterized in that, Based on the updated radial velocity and the target radial velocity, the clustering algorithm is used to determine the aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster, including: Based on the target point cloud coordinates, target point cloud azimuth difference, updated radial velocity, target radial velocity, mapped point cloud coordinates and mapped point cloud azimuth difference of the mapped previous point cloud cluster, determine the aggregate point cloud distance function between the target point cloud cluster and the mapped previous point cloud cluster. Based on the aggregated point cloud distance function, the clustering algorithm is used to determine the aggregated point cloud cluster of the target point cloud cluster and the mapped previous point cloud cluster.

6. The method according to claim 1, characterized in that, Before determining the mapping relationship between the target point cloud cluster and the previous point cloud cluster based on the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster, the method further includes: The vehicle is set up with a relative coordinate system based on the millimeter-wave radar device. The average point cloud coordinates of the target point cloud cluster in the relative coordinate system are taken as the target point cloud coordinates, and the average point cloud coordinates of the previous point cloud cluster in the relative coordinate system are taken as the previous point cloud coordinates. The obstacle running speed of the target obstacle is determined based on the target point cloud coordinates and the previous point cloud coordinates; Based on the relative coordinate system, and according to the obstacle's running speed, the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster are determined respectively.

7. The method according to claim 6, characterized in that, Based on the relative coordinate system, and according to the obstacle's running speed, the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster are determined, including: Determine the first connecting line between the origin of the relative coordinate system and the center of the target point cloud, the second connecting line between the origin of the coordinate system and the previous point cloud cluster, and the third connecting line between the center of the target point cloud cluster and the center of the previous point cloud cluster, and determine the first operating azimuth angle between the first connecting line and the third connecting line, and the second operating azimuth angle between the second connecting line and the third connecting line; Based on the obstacle's running speed and the first running azimuth angle, determine the target radial velocity corresponding to the target point cloud cluster; Based on the obstacle's running speed and the second running azimuth angle, determine the previous radial velocity corresponding to the previous point cloud cluster.

8. An obstacle detection device, characterized in that, include: The point cloud cluster determination module is used to acquire the target frame point cloud obtained when the millimeter-wave radar detects the target obstacle, and to use a clustering algorithm to determine the target point cloud cluster of the target frame point cloud, as well as the previous point cloud cluster of the previous frame point cloud of the target frame point cloud. The mapping relationship determination module is used to determine the mapping relationship between the target point cloud cluster and the previous point cloud cluster based on the target radial velocity corresponding to the target point cloud cluster and the previous radial velocity corresponding to the previous point cloud cluster. The point cloud cluster mapping module is used to map the previous point cloud cluster to the target point cloud cluster according to the mapping relationship, and to determine the updated radial velocity corresponding to the mapped previous point cloud cluster. The obstacle information determination module is used to determine the target point cloud cluster and the aggregated point cloud cluster of the mapped previous point cloud cluster based on the updated radial velocity and the target radial velocity using the clustering algorithm, and to determine the obstacle information of the target obstacle based on the aggregated point cloud cluster.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the obstacle detection method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the obstacle detection method according to any one of claims 1-7.

Images