In order to improve the inspection efficiency for needle selectors, a set of automatic needle selector inspection system based on image processing was designed. The system was composed of an ordinary industrial camera, a needle selector controller, a needle selector placement platform and a personal computer. The needle selector is checked by the gray value of the knife head at the two extreme positions. By analyzing the swing rule of the needle selector knife head, the feasibility of using the gray value of the needle selector knife head to judge the knife head swing operation was verified. The Python software was adopted to achieve image acquisition and cropping, and gray value of each cutter head was efficiently extracted by combine Otsu, contour detection, image erosion and other algorithms. The results were compared with the pre-solved threshold to achieve normal operation of the cutter head measurement and judgment. The designed human-computer interaction interface can display the inspection status in real time, and it saves the pictures of the error frames of the needle selector in the designated directory and the error log in the designated file for reference and analysis. The completed system has low cost and no special experimental environment requirements. The actual test of the needle selector that simulates error indicates that the inspection system can effectively realize the detection of the needle selector blade.
万方数据