With the rapid development of electronic manufacturing industry in our country, PCB circuit defects detection is particularly important in order to improve the Printed Circuit Board (PCB) quality and reduce the cost of production. In many detection methods, Automatic Optical Inspection (AOI) is the champion with its automation, high speed and high resolution detection ability.This paper builds Automatic Optical Inspection system hardware platform by using computer vision technology and virtual instruments technique; and realizes the movement control of the X-Y electronic translation table. Combining with the characteristics of the actual tested image, the image preprocessing, recognition and matching are applied to the PCB of common defects detection. Based on the analysis of the detection speed factors, this paper presents a path optimization algorithm which detects components fast.This algorithm can improve efficiency and satisfy the real-time requirement.The results indicate that this system gets good detection effect for the common defects, and has a certain reference value for other computer vision system development. |