An Adaptive Corner Extraction Method of Point Cloud for Machine Vision Measuring System

The corner extraction of point cloud is an important part of product quality testing in the field of machine vision, the key problem of corner extraction is curvature estimation, so in this paper, an adaptive corner extraction method of point cloud is given. In this method, firstly make use of the image edge extraction method to obtain the edge points, and sorting the points to obtain the ordered edge points, secondly using the adaptive method to calculate the support region for the straight line fitting, in the end, from the intersection of the straight lines to get the curvature angle. Using this method can obtain the corners and borders of point cloud, and solve the problem of support region size when calculating curvature or fitting straight line. The experimental results show that this method has good anti-interference, and good ability to detect and locate the corner points.