Research on Ultraviolet Image Defect Recognition System of Transmission Lines Based on Deep Learning

An intelligent management system for UAV inspection information based on deep convolutional neural network is proposed. It can be well adapted to the needs of UAV operation and management. Aiming at the convenience of UAV to inspect transmission lines, a 3D intelligent inspection technology architecture based on UAV image is proposed. A deep convolutional neural network is used to preprocess many images, videos and other information collected during UAV inspection. The fault parts in the image are marked and classified, and the detection results are fed back to the corresponding fault report. 20 targets were identified with about 85% accuracy at the same time. In the general transmission device, its accuracy can reach 75%. Experimental results show that the proposed method can be used to detect transmission line faults. This new detection method can not only effectively improve the work efficiency of UAV inspection, but also reduce the use of human resources.