In the process of manual detection of Automobile Electrical Wiring Terminal(AEWT) defects in industrial production, there are problems such as low detection efficiency, time-consuming and labor-intensive. This paper puts an online detection algorithm forwards for the circularity of AEWT based on machine vision. Firstly, video images were obtained, and the key frame was extracted by the inter-frame difference method; Then, the adaptive Canny operator was used to extract the edges of images in different color channels fused with the AND operation; finally, the circularity of the AEWT is calculated employed by the edge image, and the defective products are judged according to the calculation result. Experiments show that the online detection algorithm that we proposed for the circularity of the AWET can effectively distinguish the defective products, the detection time is less than 30ms, and the accuracy is higher than 98%, which can meet the needs of real-time detection in practical application scenarios.

No abstract

No abstract

No abstract