In order to improve the inspection success rate and velocity of the in-line Automated Optical Inspection (AOI) system of Printed Circuit Boards (PCBs)

the solder joints of a PCB are studied. The solder joint images are acquired by a structure illuminator and a 3-CCD color camera. Based on the images

the area features of the conventional kinds of the PCB solder joints such as good

excessive

poor

pseudo are extracted with respect to the key sub-region in a solder joint. Five kinds of feature matrixes models of solder joints are presented. A pattern matching algorithm for inspecting the solder joint is developed by maximum the similarity of the same kind of solder joins. To solve the problem of the thresholds determined by experience

a parameter self-adaptive learning strategy is presented. Finally

a 1040 chip solder joints PCB is inspected. Experimental results show that the success rate is up to 96.5% and the inspection time is 9 seconds using the proposed algorithm

which indicates that the proposed algorithm can achieve both high success rate and inspecting velocity.