Automatic optical inspection for detecting defective solders on printed circuit boards

Abstract

Automated systems based on optical inspection are widely used in the production of electronic systems, mainly on the printed circuit board level. Detection of defects in the early stages of production, reduce costs by lowering the number of scraps and elevate overall quality standards. One of the main areas where defects arise, are in the solders of the components on printed circuit boards. This paper studies an implementation of an automated optical inspection system for detection of defective solder in through-hole components. The solution implemented in this system, relies on the correct thresholding of the acquired image and afterwards particle analysis on a rule-based approach, instead of using techniques such as template matching to compare the solders to a known template. The proposed solution reduces the overall processing time and makes for a more flexible solution. The experimental results show the efficiency of the proposed approach with a high rate of success using a minimum cycle time.