An automated system for design-rule-based visual inspection of printed circuit boards

Abstract

The design and the implementation of an automated, design-rule-based, visual printed circuit board (PCB) inspection system are presented. The system employs mathematical-morphology-based image processing algorithms. This system detects PCB defects related to the conducting structures on PCBs by checking a set of geometric design rules. For this purpose, an image segmentation algorithm and a defect detection algorithm are designed. The defect detection algorithm is capable of verifying the minimum conductor spacing, minimum conductor trace width, and the minimum land width requirements on digital binary PCB images. Also, an existing defect detection algorithm is modified for its implementation in the system.