CN-117132879-B - Dynamic obstacle recognition method and device, storage medium and electronic device

CN117132879BCN 117132879 BCN117132879 BCN 117132879BCN-117132879-B

Abstract

The application provides a method and a device for identifying dynamic obstacles, a storage medium and an electronic device, wherein the method comprises the steps of continuously collecting point clouds of an acquisition area of a target sensor through the target sensor on cleaning equipment to obtain multi-frame point clouds; mapping points with the height above the ground and less than or equal to a first height threshold value into one image according to each frame of point cloud of the multi-frame point cloud to obtain a plurality of partial images, determining obstacles contained in each partial image by performing clustering operation on the points in each partial image in the plurality of partial images, and sequentially performing obstacle recognition on the obstacles contained in each partial image to obtain dynamic obstacles in the acquisition area. The application solves the problem that the timeliness of dynamic obstacle recognition is poor due to the large data quantity required to be processed in the dynamic obstacle recognition mode in the related technology.

Inventors

CHENG LIYE
SUN YINGRI
ZHU CHENYANG

Assignees

追觅创新科技(苏州)有限公司

Dates

Publication Date: 20260508
Application Date: 20220520

Claims (8)

1. A method of identifying a dynamic obstacle, comprising: Continuously collecting point clouds in a collecting area of a target sensor through the target sensor on the cleaning equipment to obtain multi-frame point clouds; Mapping points with the height above the ground and less than or equal to a first height threshold value into one graph according to each frame of point cloud of the multi-frame point cloud to obtain a plurality of partial graphs, wherein the first height threshold value is greater than the height of the cleaning equipment; Determining obstacles contained in each of the plurality of partial graphs by performing a clustering operation on points within each partial graph; Sequentially identifying the obstacles contained in each partial graph to obtain dynamic obstacles in the acquisition area, Mapping the points with the height above the ground and less than or equal to the first height threshold value into one graph according to each frame of the multi-frame point cloud to obtain a plurality of partial graphs, wherein the mapping comprises the following steps: the following steps are executed to each frame of point cloud of the multi-frame point cloud to obtain the multiple partial graphs, wherein when the following steps are executed, each frame of point cloud is the current frame of point cloud: Performing ground detection on the current frame point cloud to obtain a ground point cloud in the current frame point cloud; Mapping points with heights above the ground point cloud in the current frame point cloud and smaller than or equal to the first height threshold value in the current frame point cloud to a map to obtain a local map corresponding to the current frame point cloud, Performing ground detection on the current frame point cloud to obtain a ground point cloud in the current frame point cloud, including: mapping points with the height smaller than or equal to a second height threshold value in the current frame point cloud into a graph to obtain a reference graph corresponding to the current frame point cloud; calculating the height difference between any two adjacent points in the reference graph; determining two adjacent points with the height difference smaller than or equal to a height difference threshold value as candidate ground points corresponding to the current frame point cloud; clustering operation is carried out on the candidate ground points, so that a plurality of candidate ground point clouds are obtained; and determining the ground point cloud in the current frame point cloud from the plurality of candidate ground point clouds according to the point cloud parameters of the plurality of candidate ground point clouds.
2. The method according to claim 1, wherein the continuously performing, by the target sensor on the cleaning device, the point cloud acquisition on the acquisition area of the target sensor to obtain a multi-frame point cloud includes: and continuously collecting point clouds in an acquisition area of the area array TOF sensor through the area array TOF sensor to obtain the multi-frame point clouds.
3. The method of claim 1, wherein the determining the obstacles contained in each of the plurality of partial graphs by performing a clustering operation on points within each of the partial graphs comprises: Executing the following steps on each partial graph in the plurality of partial graphs to obtain an obstacle contained in each partial graph, wherein when executing the following steps, each partial graph is a current partial graph: Performing clustering operation on points in the current local graph to obtain a plurality of candidate barriers, wherein each candidate barrier of the plurality of candidate barriers corresponds to one cluster obtained by clustering; and selecting an obstacle with a size larger than or equal to a target size threshold from the plurality of candidate obstacles, and obtaining the obstacle contained in the current partial graph.
4. A method according to any one of claims 1 to 3, wherein the sequentially identifying the obstacles contained in each partial graph to obtain the dynamic obstacle in the acquisition region comprises: Obtaining the position of any obstacle in the acquisition area in the plurality of partial images by performing obstacle matching on the obstacle contained in each partial image; and determining a dynamic identification result of any obstacle according to the positions of the any obstacle in the plurality of partial graphs.
5. The method of claim 4, wherein determining the target recognition result of the any obstacle according to the positions of the any obstacle in the plurality of partial figures comprises: Determining the equipment position of the cleaning equipment according to the movement parameters of the cleaning equipment under the condition that the cleaning equipment is in a moving state; Converting the position of any obstacle in the plurality of partial figures into a position under a world coordinate system according to the equipment position of the cleaning equipment to obtain a group of position sequences of any obstacle; And determining the dynamic identification result of any obstacle according to the distance between two adjacent positions in the group of position sequences.
6. A dynamic obstacle recognition device, comprising: The acquisition unit is used for continuously carrying out point cloud acquisition on an acquisition area of the target sensor through the target sensor on the cleaning equipment to obtain multi-frame point cloud; The mapping unit is used for mapping points with the height above the ground and smaller than or equal to a first height threshold value into one map according to each frame of point cloud of the multi-frame point cloud so as to obtain a plurality of partial maps, wherein the first height threshold value is larger than the height of the cleaning equipment; A clustering unit configured to determine an obstacle included in each of the partial graphs by performing a clustering operation on points within each of the partial graphs; An identification unit for sequentially identifying the obstacles contained in each partial graph to obtain dynamic obstacles in the acquisition area, Mapping the points with the height above the ground and less than or equal to the first height threshold value into one graph according to each frame of the multi-frame point cloud to obtain a plurality of partial graphs, wherein the mapping comprises the following steps: the following steps are executed to each frame of point cloud of the multi-frame point cloud to obtain the multiple partial graphs, wherein when the following steps are executed, each frame of point cloud is the current frame of point cloud: Performing ground detection on the current frame point cloud to obtain a ground point cloud in the current frame point cloud; Mapping points with heights above the ground point cloud in the current frame point cloud and smaller than or equal to the first height threshold value in the current frame point cloud to a map to obtain a local map corresponding to the current frame point cloud, Performing ground detection on the current frame point cloud to obtain a ground point cloud in the current frame point cloud, including: mapping points with the height smaller than or equal to a second height threshold value in the current frame point cloud into a graph to obtain a reference graph corresponding to the current frame point cloud; calculating the height difference between any two adjacent points in the reference graph; determining two adjacent points with the height difference smaller than or equal to a height difference threshold value as candidate ground points corresponding to the current frame point cloud; clustering operation is carried out on the candidate ground points, so that a plurality of candidate ground point clouds are obtained; and determining the ground point cloud in the current frame point cloud from the plurality of candidate ground point clouds according to the point cloud parameters of the plurality of candidate ground point clouds.
7. A computer-readable storage medium, characterized in that the computer-readable storage medium comprises a stored program, wherein the program when run performs the method of any one of claims 1 to 5.
8. An electronic device comprising a memory and a processor, wherein the memory has stored therein a computer program, the processor being arranged to perform the method of any of claims 1 to 5 by means of the computer program.

Description

Dynamic obstacle recognition method and device, storage medium and electronic device [ Field of technology ] The application relates to the field of smart home, in particular to a method and a device for identifying dynamic obstacles, a storage medium and an electronic device. [ Background Art ] During operation of the cleaning device, dynamic obstacle detection may be performed by sensors configured on the cleaning device to identify obstacles that are moving or moving away from the cleaning device. After the dynamic obstacle is identified, the obstacle avoidance can be performed through the configured obstacle avoidance strategy, so that abnormal conditions such as equipment damage and the like caused by collision with the dynamic obstacle are avoided. Currently, when dynamic obstacle recognition is performed, the acquired multi-frame images are used for obstacle recognition, and the dynamic obstacle is recognized according to the position change of the obstacle in the multi-frame images. However, in the above-mentioned dynamic obstacle recognition method, the dynamic obstacle cannot be recognized in time due to the large amount of data to be processed, so that the situation that the cleaning device collides with the dynamic obstacle still occurs. As can be seen from the above, the dynamic obstacle recognition method in the related art has a problem of poor timeliness of dynamic obstacle recognition due to a large amount of data to be processed. [ Invention ] The application aims to provide a method and a device for identifying dynamic obstacles, a storage medium and an electronic device, which at least solve the problem that the timeliness of dynamic obstacle identification is poor due to the large data volume required to be processed in the identification mode of the dynamic obstacles in the related technology. The application aims at realizing the following technical scheme: according to one aspect of the embodiment of the application, a dynamic obstacle identification method is provided, which comprises the steps of continuously carrying out point cloud acquisition on an acquisition area of a target sensor through the target sensor on cleaning equipment to obtain multi-frame point clouds, mapping points with heights above the ground and smaller than or equal to a first height threshold value into one graph according to each frame of the multi-frame point clouds to obtain a plurality of partial graphs, determining obstacles contained in each partial graph through clustering operation on the points in each partial graph, and sequentially carrying out obstacle identification on the obstacles contained in each partial graph to obtain the dynamic obstacle in the acquisition area. In an exemplary embodiment, the continuously performing, by the target sensor on the cleaning device, the point cloud acquisition on the acquisition area of the target sensor to obtain a multi-frame point cloud includes continuously performing, by the area array ToF sensor, the point cloud acquisition on the acquisition area of the area array ToF sensor to obtain the multi-frame point cloud. In an exemplary embodiment, the mapping the points with the height above the ground and less than or equal to the first height threshold into one map according to each frame of the multi-frame point cloud to obtain a plurality of local maps includes performing the following steps on each frame of the multi-frame point cloud to obtain the plurality of local maps, wherein when performing the following steps, each frame of point cloud is a current frame of point cloud, ground detection is performed on the current frame of point cloud to obtain a ground point cloud in the current frame of point cloud, and the points with the height above the ground point cloud in the current frame of point cloud and less than or equal to the first height threshold in one map are mapped into one map to obtain one local map corresponding to the current frame of point cloud. In an exemplary embodiment, the ground detection is performed on the current frame point cloud to obtain a ground point cloud in the current frame point cloud, and the ground detection comprises the steps of mapping points with heights smaller than or equal to a second height threshold value in the current frame point cloud into a graph to obtain a reference graph corresponding to the current frame point cloud, calculating a height difference between any two adjacent points in the reference graph, determining two adjacent points with the height difference smaller than or equal to a height difference threshold value as a set of candidate ground points corresponding to the current frame point cloud, performing clustering operation on the set of candidate ground points to obtain a set of multiple candidate ground point clouds, and determining the ground point cloud in the current frame point cloud from the set of multiple candidate ground point clouds according to a point cloud parameter of each can