Study on Curved Surface Fitting of Countersink Based on Point Cloud

Abstract

According to high precision of point cloud, countersink quality detection based on point cloud is an effective way to detect defect of countersink. One of the major processes in countersink quality detection is curved surface fitting. In this study, we eliminate point cloud interference considering curved surface which is not absolutely smooth and propose a curved surface fitting algorithm for building cone model by using circular curves and conical surfaces. The fitting algorithm is designed to establish model precisely considering diameter of edge, depth, and nest angel of countersink. In addition, to validate the performance of the algorithm, a computer vision approach which is developed in literature is presented and its results are compared with those obtained by using the fitting algorithm based on point cloud. Experimental results show that the fitting algorithm can be regard as an effective and precise algorithm for curved surface fitting under external interference in terms of experimental results. Furthermore, this study reveals that the proposed algorithm is more precision and has better anti-noise performance compared with computer vision detection approach.

Keywords

• Countersink
• computer vision
• curved surface fitting
• feature extraction of point cloud
• RANSAC