Research on SMT solder joint image segmentation

Research on SMT solder joint image segmentation can formed feedback information of the products assembly which is used to control and adjust the SMT assembly, and is useful to guide SMT to improve the quality and process of SMT products. In our study, we improve a novel approach of SMT solder joint image segmentation. During our method, the process has three steps. At first, the RGB image of solder joint is converted into HSV image. Second, we segment the solder joint by color domain 2D histogram threshold. Last, mathematics morphology processing is used to improve the segment accuracy. Experimental results show that the proposed method retains the structure of the solder joint images leading to an effective segmentation.