A Novel Method for Detecting Thickness Defects in Metal Components Based on Point Cloud and Model Registration

This paper presents a novel method for detecting thickness defects in metal parts. Firstly, a three-dimensional point cloud acquisition system is established to acquire point cloud information of the inspected part. Defect detection on the part is achieved through operations such as denoising, registration, and feature extraction on the point cloud data. The proposed method optimizes the point-to-point nearest neighbor iterative algorithm based on the registration of point cloud with an STL model. An empirical exploration delves into the intricacies of a metallic component, scrutinizing minute thickness fluctuations of up to 0.1 mm. Through rigorous experimentation, the viability and efficiency of the innovative method are vividly showcased, highlighting its tangible impact on practical applications.