Micro-drilling End Machine Vision-based Pcb Defect Detection Study

As important foundation of electronic system, printed circuit board (PCB) plays a key role in the electronic information industry. In the flow of PCB production, hole forming is an important processing procedure which determines the final quality of PCB, and the blade defects of PCB micro drill bit is the determinant factor of drill hole. At present, for defects of micro drill bit blade surface,it's adopted by the method of selective check using contact microscope. The inspecting precision is not good and it's hard to control the whole quality's level.This dissertation focuses on the solution of the above problems. Based on the principle of digital image processing, innovative and exploring research on the algorithms of the blade surface defects inspection for PCB micro drill bit is presented. The main contents are as follows:1. Based on the shape characters of micro drill bit,the optical imaging system, especially the illumination system is designed. With the specially designed LED ring light source, the high-contrast steady image of testing target has been obtained, and a foundation has been laid for the following image processing.2. The category defects of the first blade face of PCB micro drill bit blade are analyzed. The whole algorithm flow of defects inspection including image preprocessing, image segmentation, image feature extraction, characteristics analysis and the comprehensive judgment of results, has been made.3. The image processing algorithm of defects inspection is analyzed synthetically in all its the aspects such as segmentation, filter, enhancement, mark, edge tracking, extraction of characteristic points and parameter calculation. And the method of utilization of the Laplace of Gauss filter and image segmentation in micro drill bit edge extraction has been proposed. The complete flow of micro drill bit's defects inspection has been realized and reliable results have been obtained.4. The improvement of the system is proposed in the following aspects: optical parameters improvement, man-machine interface optimization, stability of algorithm of calculation speed further optimization and whole system automation. The most of important contribution is the utilization of the Laplace of Gauss filter and the image segmentation in micro drill bit edge extraction. It is the first domestic report.The research results have been used in commercial products and the appraisal of customer is affirmative.