The work aims to design the chip resistance direction detection system, in order to solve the problems such as low detection efficiency and low detection precision in the process of ordering the current chip resistance. With the machine vision software Halcon as the platform, the improved shape-based template matching algorithm was used to carry out the edge detection and filtering based on the Canny edge detection operator and Sobel operator, so as to generate the edge point position vector and gradient vector of the template image. The template matching was conducted with the inner product of the template edge point and the corresponding point in the image search area. The recognition time of the detection system was 33 ms, and its average recognition accuracy was more than 99.5%, which was better than that of the traditional manual detection. With good speed and accuracy, the system can meet the requirements of industrial production.