CN-115359453-B - Road surface obstacle detection method and system based on three-dimensional point cloud

CN115359453BCN 115359453 BCN115359453 BCN 115359453BCN-115359453-B

Abstract

The invention relates to a road surface obstacle detection method and system based on three-dimensional point clouds, which are used for carrying out spatial clustering based on density on road surface three-dimensional point cloud data acquired by a four-eye camera to obtain a plurality of clusters, obtaining the width of an obstacle by calculating the difference between the maximum value and the minimum value of x coordinates of point clouds in the clusters under the three-dimensional coordinate axes, obtaining the height of the obstacle by calculating the difference between the maximum value and the minimum value of y coordinates of the point clouds in the clusters, and obtaining the distance between a body and the obstacle by calculating the mean value of z coordinate values of points clouds in the clusters. The method is directly realized through three-dimensional point cloud data generated by the four-eye camera, so that the perception module of the unmanned system obtains depth information and can be in one-to-one correspondence with the physical coordinates of the three-dimensional space. The method is applied to the perception of the road surface obstacle in the unmanned system, so that the existence of the obstacle can be detected, and the size of the obstacle and the distance between the obstacle and the unmanned vehicle can be detected.

Inventors

Wang Caiju
LI XIAOMAO
CAO LIANG
PENG YAN
XIE SHAORONG
WU YIQIANG
ZHU YUHE

Assignees

上海大学
上海大学

Dates

Publication Date: 20260421
Application Date: 20220812
Priority Date: 20220812

Claims (5)

1. The road obstacle detection method based on the three-dimensional point cloud is characterized by comprising the following steps of: carrying out spatial clustering based on density on target point cloud data to obtain a plurality of clusters, wherein the target point cloud data is obtained by shooting a road scene by a four-eye camera; Calculating the difference between the maximum value and the minimum value of the x coordinates of the point cloud in the clusters under a three-dimensional coordinate system aiming at each cluster to obtain the width of the obstacle, wherein the three-dimensional coordinate system is a coordinate system established by taking a vehicle as an origin and the advancing direction of the vehicle as a z axis; calculating the difference between the maximum value and the minimum value of the y coordinates of the point cloud in the cluster to obtain the height of the obstacle; Calculating the mean value of the z coordinate values of each point cloud in the cluster to obtain the distance between the vehicle and the obstacle; before the spatial clustering of the target point cloud data based on density, the method further comprises the steps of preprocessing three-dimensional point cloud data acquired by the four-eye camera: Setting a first threshold value along the x-axis direction by adopting a threshold value segmentation method, filtering point clouds with the x-axis direction being larger than the first threshold value, and realizing the effect of filtering noise points on two sides of a road surface; downsampling the extracted point cloud data of the region of interest to obtain downsampled point cloud data; Performing outlier filtering on the downsampled point cloud data, specifically including: calculating the average distance between each downsampling point and k nearest neighbor points to the downsampling point to obtain the average distance corresponding to each downsampling point; acquiring the average value and standard deviation of the average distance; acquiring noise data according to the average distance, the average value, the standard deviation and a set threshold value; filtering the noise data from the down-sampling point cloud data; the obtaining noise data according to the average distance, the average value, the standard deviation and the set threshold value specifically includes: for each point p in the downsampled point cloud data, judging whether the point p meets a formula or not If yes, determining the point p as a noise point, otherwise, determining the point p as not the noise point; Wherein, the Representing the average distance of point p from all points in the k neighborhood, The standard deviation of the distribution is indicated, Represents the mean of the distribution, and t represents the threshold.
2. The method according to claim 1, wherein the downsampling the extracted point cloud data of the region of interest to obtain downsampled point cloud data, specifically includes: dividing the point cloud data of the region of interest into a plurality of cube grids; selecting a non-empty cube grid in the cube grids; Calculating the mass centers of all points in the non-empty cube grids aiming at each non-empty cube grid to obtain mass center points; And replacing all points in the non-empty cube grid with the centroid points to obtain downsampling points corresponding to the non-empty cube grid.
3. The method according to claim 1, wherein the spatial clustering of the target point cloud data based on density specifically comprises: acquiring point cloud sets to be processed The point cloud set to be processed is composed of the three-dimensional point cloud data; Initializing a set of core objects Number of clusters Non-accessed sample set Clustering of clusters ; Traversing the sample set S according to the formula And Obtaining points in the core object set ; Representation points And R represents the radial length of the neighborhood; represented in sample set S in dots A sample set consisting of points in a circle with radius r and a center, wherein MinPts represents the number of the least points contained in one cluster; representing the total number of the set elements; judging whether the core object set is an empty set or not to obtain a first judging result, ending if the first judging result is yes, otherwise, randomly selecting a core object from the core object set And initializing the current cluster core object set Number of clusters Initializing a current cluster sample set Updating a set of unvisited samples ; Judging whether the core object set of the current cluster is an empty set or not to obtain a second judging result, if not, updating the core object set Otherwise, the current cluster is clustered Added to the cluster set C and according to Updating the core object set, and returning to the step of judging whether the core object set is an empty set; from the current cluster core object set A core object is fetched ' Acquiring and said ' Neighborhood sample set with European distance less than or equal to E Order-making Updating a current cluster sample set Updating a set of unvisited samples Updating Step "determine whether the current cluster core object set is an empty set".
4. A three-dimensional point cloud-based road obstacle detection system, comprising: The preprocessing module is used for preprocessing the three-dimensional point cloud data acquired by the four-eye camera before the density-based spatial clustering of the target point cloud data; The system comprises a clustering model, a target point cloud data acquisition module and a target point cloud data acquisition module, wherein the clustering model is used for carrying out density-based spatial clustering on the target point cloud data to obtain a plurality of clusters; The obstacle width calculation module is used for calculating the difference between the maximum value and the minimum value of the x coordinate of the point cloud in the cluster under the three-dimensional coordinate system aiming at each cluster to obtain the width of the obstacle, wherein the three-dimensional coordinate system is a coordinate system established by taking a vehicle as an origin and the advancing direction of the vehicle as a z axis; The obstacle height calculation module is used for calculating the difference between the maximum value and the minimum value of the y coordinates of the point clouds in the clusters to obtain the height of the obstacle; The distance calculation module is used for calculating the mean value of the z coordinate values of each point cloud in the cluster to obtain the distance between the body and the obstacle; The preprocessing module specifically comprises: The interest region extraction submodule is used for extracting the interest region from the road surface three-dimensional point cloud data; setting a first threshold along the x-axis direction, filtering point clouds with the x-axis direction being larger than the first threshold, and realizing the effect of filtering noise points on two sides of a road surface; The downsampling submodule is used for downsampling the extracted point cloud data of the region of interest to obtain downsampled point cloud data; The outlier filtering sub-module is used for performing outlier filtering on the down-sampling point cloud data, and specifically comprises the following steps: calculating the average distance between each downsampling point and k nearest neighbor points to the downsampling point to obtain the average distance corresponding to each downsampling point; acquiring the average value and standard deviation of the average distance; acquiring noise data according to the average distance, the average value, the standard deviation and a set threshold value; filtering the noise data from the down-sampling point cloud data; the obtaining noise data according to the average distance, the average value, the standard deviation and the set threshold value specifically includes: for each point p in the downsampled point cloud data, judging whether the point p meets a formula or not If yes, determining the point p as a noise point, otherwise, determining the point p as not the noise point; Wherein, the Representing the average distance of point p from all points in the k neighborhood, The standard deviation of the distribution is indicated, Represents the mean of the distribution, and t represents the threshold.
5. The system according to claim 4, characterized in that said downsampling sub-module comprises in particular: the dividing subunit is used for dividing the point cloud data of the region of interest into a plurality of cube grids; a non-empty cube grid selecting subunit, configured to select a non-empty cube grid in the cube grids; A centroid point calculating subunit, configured to calculate, for each non-empty cube grid, a centroid of all points in the non-empty cube grid, to obtain a centroid point; and the downsampling subunit is used for replacing all points in the non-empty cube grid by using the centroid point to obtain downsampling points corresponding to the non-empty cube grid.

Description

Road surface obstacle detection method and system based on three-dimensional point cloud Technical Field The invention relates to the field of three-dimensional point cloud processing technology and computer vision, in particular to a road surface obstacle detection method and system based on three-dimensional point cloud. Background The refinement of the unmanned technology has remarkable research value and social value for the whole automobile industry, and not only avoids traffic accidents caused by drunk driving, fatigue driving and the like, thereby improving the safety, but also reduces traffic jams caused by overtaking, jam adding and the like. A complete unmanned system mainly comprises three modules of perception, decision and control, wherein the perception module is the basis for realizing functions of other modules, and target detection is an important task of the perception module and is used for acquiring position and type information (such as vehicles, pedestrians, roadblocks and the like) of objects in a space, so that the system has important guiding effects on subsequent path planning, collision avoidance and the like. The object detection technology is one of three tasks in the field of computer vision, which aims to acquire position and category information of an object. The road obstacle target detection is to further limit the range based on the target detection, firstly, the scene is set as the road on which the vehicle runs, and secondly, the targets are pedestrians, vehicles, roadblocks and the like on the road surface. Road obstacle target detection requires that all obstacles be found out in the acquired data, and located and classified. In the traditional road surface obstacle detection method, if only two-dimensional image data are subjected to target detection, although targets in the images can be accurately identified, the depth information which is essential to a perception module is lacking when the targets are mapped to a three-dimensional space, namely, an unmanned vehicle perceives the existence of an obstacle, but never knows the distance between the obstacle and the unmanned vehicle and the obstacle size, so that due help cannot be provided for later path planning and obstacle avoidance. Based on this, there is a need for a road obstacle detection method and system that can learn the distance between the obstacle and itself and the obstacle size. Disclosure of Invention The invention aims to provide a road surface obstacle detection method and system based on three-dimensional point cloud, which are applied to the perception of road surface obstacles in an unmanned system, and can detect the existence of the obstacles, the size of the obstacles and the distance between the obstacles and the unmanned vehicle, so that the traffic accidents of the unmanned vehicle are reduced to a great extent, and the safety is improved. In order to achieve the above object, the present invention provides the following solutions: a road obstacle detection method based on three-dimensional point cloud, comprising: carrying out spatial clustering based on density on target point cloud data to obtain a plurality of clusters, wherein the target point cloud data is obtained by shooting a road scene by a four-eye camera; Calculating the difference between the maximum value and the minimum value of the x coordinates of the point cloud in the clusters under a three-dimensional coordinate system aiming at each cluster to obtain the width of the obstacle, wherein the three-dimensional coordinate system is a coordinate system established by taking a vehicle as an origin and the advancing direction of the vehicle as a z axis; calculating the difference between the maximum value and the minimum value of the y coordinates of the point cloud in the cluster to obtain the height of the obstacle; and calculating the mean value of the z coordinate values of each point cloud in the cluster to obtain the distance between the vehicle and the obstacle. As an optional implementation manner, before the density-based spatial clustering is performed on the target point cloud data, the method further comprises the steps of preprocessing three-dimensional point cloud data acquired by the four-eye camera: Extracting an interested region from the three-dimensional point cloud data; downsampling the extracted point cloud data of the region of interest to obtain downsampled point cloud data; and performing outlier filtering on the downsampled point cloud data. As an optional implementation manner, the downsampling the extracted point cloud data of the region of interest to obtain downsampled point cloud data specifically includes: dividing the point cloud data of the region of interest into a plurality of cube grids; selecting a non-empty cube grid in the cube grids; Calculating the mass centers of all points in the non-empty cube grids aiming at each non-empty cube grid to obtain mass center points; And replacing all p