Vehicle Target Intelligent Recognition Method Based on Point Cloud Multi-feature Classification
LiDAR is one of the important sensors of intelligent vehicle environment perception.Aiming at the prob-lem of poor effect of laser point cloud perception and vehicle recognition,an intelligent vehicle target recognition method based on multi-feature classification of point cloud is proposed.Firstly,ground segmentation and spatial clustering are used to preprocess point cloud data,and several targets to be identified are obtained.Secondly,for the target to be recognized,the multi-feature classification is constructed from the dimensions of vehicles,body materials and point cloud features.Finally,a training set is made to train the support vector machine and the trained classifier is used to recognize the target.Six scenarios from the Baidu Apollo dataset are used in the experi-ment.The accuracy of the proposed method is above 90%,and the recognition effect is improved compared with that of the existing methods.
万方数据
维普
