Application of machine vision in capacitor appearance defect detection

Abstract

The traditional capacitor appearance defect detection adopts manual detection, which has low efficiency, high error rate and high cost. In order to overcome the shortcomings of manual detection and improve the automation of capacitor production, a machine vision based capacitor defect detection system is designed. Firstly, the image is collected, pre-processed, and matched to the capacitor area. Then the threshold segmentation is used to detect the defects of the overflow and epoxy surface pores. The template matching is used to detect the characters and the shell damage defects. Finally, to meet the testing requirements of different standards，the Blob analysis is used to extract the defect features and set the threshold parameters. According to the experimental results of the prototype, the detection system greatly improves the detection efficiency and accuracy.

Keywords

• machine vision
• capacitance appearance
• defect detection
• threshold segmentation;template matching