Vehicle Steering Gear Sleeve Defect Detection Method Based on Machine Vision

Abstract

This project takes the steering gear sleeve as the research object, and uses computer vision technology to detect it. By using this method, the Canny operator is adaptive filtered and integrated with the watershed method, and many features of the image are extracted. SVM classifier was established for three representative defects, such as lack of material, burr and coating. It is used to identify the steering gear sleeve. A test platform is set up to automatically classify and remove defective parts of automobile steering gear sleeve based on UR5 robot. The accuracy and stability of detection are improved by an improved edge segmentation algorithm. The experimental results show that the system can diagnose the fault of the axle sleeve of the automobile steering gear and cooperate with the operating arm to make it possible to automatically detect and remove the fault products.