Workpiece Point Cloud Segmentation Method Based on Feature Point Extraction
A workpiece point cloud segmentation method based on feature point extraction was pro-posed to address the issue of inaccurate segmentation of adhesive workpiece point clouds caused by the close proximity of workpieces in actual production.The random sampling consistency algorithm was used to filter out irrelevant point clouds,and feature points were extracted by using normal vector fea-tures of point clouds.The point clouds with feature points removed were segmented by using the Euclid-ean clustering algorithm,and the workpiece point cloud was finally registered and completed by using the template point cloud.The experimental results show that this method can increase the success rate of workpiece point cloud segmentation from 63.6%to 81.8%.
point cloud segmentationnormal vectorfeature point extractionpoint cloud registration
万方数据
维普
