Obstacle classification method and device, storage medium and computer equipment

Abstract

The invention relates to an obstacle classification method. The method comprises the steps: acquiring point cloud data to be processed, wherein the to-be-processed point cloud data comprises local point cloud data covered by a position area where a target obstacle is located in the original point cloud data; rasterizing the to-be-processed point cloud data to obtain a two-dimensional grid map; based on spatial points in each grid in the two-dimensional grid map, respectively determining a target statistical parameter of each grid, and obtaining a statistical feature map corresponding to the target statistical parameter; determining target image features based on the statistical feature map through a predetermined convolutional neural network, and performing classification based on the target image features to obtain confidence coefficients that the target obstacle belongs to each candidate category; and determining the category to which the target obstacle belongs from the candidate categories according to the confidence coefficients. The scheme provided by the invention is high in classification accuracy.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to an obstacle classification method, device, computer-readable storage medium and computer equipment. Background technique [0002] In recent years, more and more environmental sensing devices are used to perceive the surrounding environment, and the original point cloud data used to characterize the surrounding environment is obtained, and then based on the detection of obstacles in the surrounding environment based on the original point cloud data, obstacle detection is carried out. Classification is to determine the category to which the obstacle belongs in each candidate category. [0003] In the traditional classification method, the characteristics of the point cloud data are artificially designed, such as the three-dimensional covariance matrix and the reflection intensity probability distribution characteristics, and then through the SVM (Support Vector Machine,...

AI Technical Summary

Problems solved by technology

[0004] Based on this, it is necessary to provide an obstacle classification method, device, computer-readable storage medium and computer equipment for the technical problems of time-consuming feature extraction and low classification accuracy in traditional methods

Images