An Automated Fault Detection System for PCB Design using Digital Image Processing

In Electronics industries in order to communicate design information for the manufacturing of various types of PCBs Gerber format file is used. Since the manufacturing of PCBs is costly process so any defect during manufacturing must be avoided, especially with the high complexity level or many of today's boards. A PCBs Defect detection and classification system uses machine vision which is used to find difficulties occurred in manual inspection. It eliminates subjective aspects and gives fast, quantitative and dimensional quality. It is also very crucial to classify the defects in order to examine and identify root causes of the defects. Defects present in PCBs are as short or open circuit, under-etch, mouse-bite and missing hole. This project proposes a PCB defect detection and classification based on imaging morphological algorithm for segmentation and simple image processing theories. Some PCBs defects exists only in certain groups. Hence, it is obvious that to improve image processing algorithm for applying a segmentation exercise. The referential approach has been implemented on defected PCBs images and template in order to detect number of defects present on bare PCB just before the etching process, because etching generally contributes most damages found on PCBs. So, PCB inspection system is then improved by using a geometrical image registration, minimum Thresholding and median filtering in order to solve alignment and uneven illumination problem.