Obstacle recognition method based on 3D point cloud data and computer equipment

Abstract

The invention designs an obstacle recognition method based on 3D point cloud data and computer equipment. The method comprises the steps of obtaining 3D point cloud data, and mapping the 3D point cloud data to a plane to obtain a grid map; obtaining an obstacle center area from the grid map by using a grid clustering method, and obtaining an obstacle reference point according to the obstacle center area; screening points in grid units outside the central area of the obstacle in the grid map by utilizing the obstacle reference point to obtain an obstacle edge area; and determining obstacle information according to the obstacle center area and the obstacle edge area. By adopting the method, the real-time performance is ensured by acquiring the central area of the obstacle through the grid clustering method, and in addition, the areas except the central area of the obstacle are clustered point by point, so that the phenomena of under-segmentation and over-segmentation are greatly reduced,and the accuracy of acquiring obstacle information is effectively improved.

Description

technical field [0001] The present application relates to the technical field of laser radar point data processing, in particular to an obstacle recognition method based on 3D point cloud data, computer equipment and storage media. Background technique [0002] With the continuous development of lidar technology, 3D lidar is used as an important environment perception sensor in unmanned vehicles. For example: use lidar to scan the surrounding environment and generate 3D point cloud data, and obtain accurate environmental information by processing the 3D point cloud data. [0003] The traditional point cloud data processing method uses the grid unit as the basic element to perform calculations. However, this method is prone to under-segmentation when the point cloud is dense; or it is prone to over-segmentation when the point cloud data is sparse; Accurate and efficient obstacle clustering also affects the accuracy of obtaining environmental information. Contents of the in...

AI Technical Summary

Problems solved by technology

[0003] The traditional point cloud data processing method uses the grid unit as the basic element to perform calculations. However, this method is prone to under-segmentation when the point cloud is dense; or it is prone to over-segmentation when the point cloud data is sparse; Accurate and efficient obstacle clustering, but also affects the accuracy of obtaining environmental information

Images