A fast computer vision system for defect detection of rubber keypad

In order to detect the defected rubber keypad the factory has to devote massive manpower and financial resources, however the efficiency and accuracy is low. In this paper, a vision-based high-speed system for the detection of rubber keypad is designed and developed. The system consists of a lighting module, an image acquisition module, a supporting frame, a computer that hosts the detection algorithm and the conveyer belt system. The design of the conveyer belt system ensures that the rubber keypad moved accordingly to the order from the computer, the lighting module ensures the high quality image could be captured. The paper also proposes a new circle target segmentation method to segment the circle rubber keypad. The system detects the defect of the rubber keypad by processing the image captured and classifies the rubber keypad. The experimental results demonstrate that the system could detect the defects of rubber keypad quickly and accurately.