Circuit board defect detection based on image processing

Abstract

Because the conventional printed circuit board defect detection is slow, a method base on the edge of image is proposed, this method can quickly locate and detect the detects on the circuit board. The algorithm combines gradient direction characteristics of image edge gray, the different characteristics of different defects in standard image and defect image are used to classify the defects, and then identify the type of defects. Experiments show that this method can detect the short circuit, opening, burr, hole and other defects of the circuit board quickly.