Abstract:In view of the two testing problems of the attachment and bending deformation of the single anode guide rod claw head, a method based on machine vision technology was proposed to detect the working condition of the steel claw of the single anode guide rod. In the scheme, two cameras were used to take pictures from different stations, then the image of the single anode pole steel claw was preprocessed through median filtering and image segmentation technology; finally, an image detection criterion based on the deformation of the claw head was developed to judge whether the steel claw to meet the requirements of normal work.The practical experiments show that the scheme can be used to detect the working condition of the single anode guide rod steel claw effectively, and the detection efficiency is greatly improved while the detection accuracy is high.