PCB defect detection system based on image processing

Under the background of intelligent manufacturing, quality inspection printed circuit board (PCB), as carrier electronic products, plays an important role. In view shortcomings existing traditional detections, this paper has put forward a PCB detection method based on machine vision, focusing image preprocessing algorithm and analysis. The median filter was used to remove noise in image. Logarithmic transformation enhance contrast. Sobel operator for edge obtain geometric center defect area. On basis, registered by template matching method, difference between preprocessed be tested standard calculated get map. Subsequently, morphological slight differences end, simulation results have shown that proposed good effect high precision detection.