BGA void detection in X-ray images based on a new blob detector

Automatic Ball Grid Array (BGA) void detection in X-ray is essential since voids can degrade the performance of the board and cause failure. In this paper, we propose a new blob detector for automatic detection of BGA voids. The proposed blob detector, which obtains candidate voids by assigning a minimum point for each blob-like region, is based on the average box filter and the image local brightness variation. Owning to the utilization of different sizes of average box filters, the proposed method is able to detect different sizes of voids. Furthermore, since the new blob detector takes advantage of the brightness variation of the local region, it is robust to the image brightness changes and ball component interferences. Experimental results show that the proposed method outperforms the other method, which detects the voids by edge detection.