PCB Defect Detection and Enhancement System Based on MATLAB

Abstract

In the process of PCB manufacturing, with the increasing demand, the complexity of IC chips, the number of pins, and the density are getting higher and higher. Therefore, the rationality of PCB layout is particularly important. In many links of PCB production and processing, the quality of solder joints needs to be tested. In recent years, in order to improve the automation and efficiency of PCB solder joint quality inspection, the automatic optical inspection system of PCB solder joint based on CCD camera image has received more and more attention. In this paper, a PCB defect detection and enhancement system based on MATLAB is studied aiming at the problem of PCB solder joint location and defect detection. The main tasks of the system include PCB surface feature image acquisition, image enhancement and segmentation preprocessing, PCB surface feature extraction, neural network model establishment and defect identification, etc., completing PCB defect location and establishing PCB defect data set, in which BP neural network is used as the judgment model of this project, It can meet the current PCB surface defect monitoring requirements.