Line defect image aided recognition method for power communication operation inspection

Aiming at the problem of low recognition accuracy caused by poor edge detection effect in line defect image auxiliary recognition method, a line defect image auxiliary recognition method for power communication operation inspection is proposed. Dark channel defogging and adaptive gamma correction are performed on the power communication operation inspection images with poor imaging affected by fog and low contrast due to illumination, so as to improve the image contrast and detail performance. The Ratio algorithm is used to detect the edge of transmission line, extract the wire from the image, establish the line defect image auxiliary recognition model based on CNN, and complete the detection of four types of defects. The test results show that the accuracy of transmission line defect identification by this method is 96.70%, which is 4.63% and 7.02% higher than that based on BP neural network and YOLOV3. The identification results can meet the needs of practical application.