Failure detection on PCBs: an image processing based approach

The concern to identify methods to obtain low cost levels in testing printed circuit boards is an important issue in quality assurance. In this paper, a fast and low cost testing system is proposed, based on images captured using optical and thermal cameras. The images are divided into blocks; all blocks are compared with reference blocks provided by reference images. Our system detects the differences and delivers a message with the name of the device or the group of devices that fail. Next, a decision tree is used, to set the trajectory of the tested board.