Considering that the traditional manual detection of micro-accessory has some problems, such as heavy workload, low efficiency and large artificial error, a kind of quality inspection system of micro-accessory has been designed. Micro-vision technology has been used to inspect quality, which optimizes the structure of the detection system. The stepper motor is used to drive the rotating micro-platform to transfer quarantine device and the microscopic vision system is applied to get graphic information of micro-accessory. The methods of image processing and pattern matching, the variable scale Sobel differential edge detection algorithm and the improved Zernike moments sub-pixel edge detection algorithm are combined in the system in order to achieve a more detailed and accurate edge of the defect detection. The grade at the edge of the complex signal can be achieved accurately by extracting through the proposed system, and then it can distinguish the qualified products and unqualified products with high precision recognition.
Camera-Based Road Damage Detection System with Edge Detection Algorithm