Defect Identification of PCB Board Based on Machine Vision

Inhalt:
With the development of semiconductor manufacturing technology, PCB industry also faces the challenge of quality problem. In this paper, aiming at the defect object of PCB board, by comparing the effect of various operators on image pretreatment, the appropriate processing operator is selected. And the algorithms of image feature extraction, recognition and matching are studied, and the intelligent detection method based on machine vision is studied. In addition, the cross-correlation image registration algorithm is used to calculate the deviation between the reference image and the first registration image through the cross-correlation function. Simulation results show that the algorithm has good registration effect and high precision in PCB defect detection.