Printed circuit board inspection using wavelet-based technique

Abstract

Malaysia is a well-known country with high quality productions in electronic manufacturing industry. One of the backbones in electronic manufacturing industry is the Printed Circuit Board (PCB) manufacturing. Current practice in PCB manufacturing requires etching. This process is irreversible. Printing, which is done before etching, caused most of the destructive defects found on the PCB. Once the laminate is etched, the defects, if exist would cause the PCB laminate to be of no use. Due to the fatigue and speed requirement, manual inspection is ineffective in inspecting every printed laminate. Therefore, manufacturers require an automated system to perform online defect detection which may occur during the printing. Hence, this project proposes an algorithm for automated visual PCB inspection that is able to automatically detect and locate any defect on a PCB laminate. Numerous approaches have been discussed in the literature so far but this work introduces a technique that is based on wavelets and multiresolution analysis. The inspection could be speed up by introducing wavelet-based algorithm for PCB defect detection and localization. The defect is detected by utilizing wavelet-based image difference algorithm. The detected defect is essential to locate the defective area on the tested PCB image. The ultimate goal of this project is to ensure the system can perform in a real-time environment with high performance.