An automated modular approach to the segmentation and inspection of printed circuit boards

This dissertation describes a new method for the inspection of printed circuit board (PCB) images. The dissertation is organized into three sections--survey of the existing PCB inspection techniques, new segmentation algorithms for the PCB images and a sub-pattern level inspection technique for the PCB images.; Section I presents the survey of the existing automated visual inspection approaches that have been reported over the last two decades. This includes a classification of the existing algorithms to date and mainly concentrates on the image analysis and fault detection strategies. A summary of the commercial PCB inspection systems is also presented towards the end of the survey.; Section II presents the new segmentation techniques developed for the segmentation of the PCB images into small windows containing primitive PCB sub-patterns. The technique proposed exploits the simplicity and the homogeneity of the PCB images and is inherently parallel. A detailed discussion of the experiments and the results obtained is also presented.; Section III presents the new inspection technique developed. The technique presented here can handle all of the defects simultaneously using a uniform approach, which is a limitation to the existing approaches. The technique presented is faster and highly parallel as it works at sub-pattern level. This section also introduces the application of neural networks and fuzzy logic in printed circuit board inspection. Experimental results which demonstrate the effectiveness of the proposed algorithms are given at the end of the paper.