The study on information extraction of BGA solders joint 2-D morphology based on wavelet modulus maximum

The shape information data of the solder joints is one of the foundations for realizing the intelligent identification of solder joint defects in the Surface Mounting Technology (SMT). In this paper, the images of BGA solder joints which collected by the automatic X-ray inspection (AXI) equipment are selected as the research object, and then two-dimensional shape information extraction technology of solder joints has been mainly studied. The basic method is as follows: the images of BGA solder joints firstly are carried out the de-noising and edge extraction processing in the transform domain after wavelet transform, then the corresponding binary images can be gained after morphological processing, lastly the useful two-dimensional shape information (perimeter, area, degree of circularity and voiding area of BGA solder joints) can be obtained by calculating with MATLB software. The study result shows that this algorithm can splendidly split the edge of BGA solder joints and their voiding to get the two-dimensional shape information data, which provides required data for achieving the intelligent identification of solder joint defects.