Research on PCB defect detection method based on principal component analysis

The surface defect detection of printed circuit board (PCB) is an important means to ensure the quality of PCB. In this paper, the method of PCB defect detection is deeply studied, and the image registration, image difference method and image segmentation in the detection process are mainly discussed. Random Hough transform is used to extract the coordinate value of the feature point and improve the accuracy of registration. In calculating the difference image, an improved method is proposed, which can effectively extract the difference between images and reduce the interference information. Based on Otsu threshold, a weighted threshold method is proposed to improve the accuracy of image segmentation. The location and size of defects in PCB images were detected by combining connected components detection and area threshold. The experimental results show that the proposed method has high accuracy in the identification of PCB surface defects and can meet the actual system requirements.