Machine Vision-Aided Quality Decision System for Solder Joint Defect Evaluation

Abstract

To improve the printed circuit board (PCB) manufacturing process, it is important to have an automatic inspection system that classifies information regarding defects in solder joints. This paper proposes a quality decision system for solder joint defect classification on a PCB. An experiment was conducted to demonstrate the application of the system. The results showed that the inspection accuracy reached 94%, which is superior to the results achieved by other methods. The results of this study provide an effective solution for the inspection of the solder joint quality.