Research on Intelligent Detection Method of Power Transmission Components Based on Image Enhancement

Transmission lines, as an important carrier of power transmission, are the part of the power grid with the widest distribution coverage. Among them, transmission components bear the important role of electrical insulation and line protection, and timely defect detection of transmission components is a key factor to ensure the stable operation of the power system. Existing defect detection of transmission components is mainly carried out by means of UAV inspection, but the image background of transmission components obtained from UAV aerial photography is complex, and the defect size is small, which is prone to problems such as misdetection and omission detection. For this reason, an image enhancement-based defect detection method for transmission components is proposed to improve the accuracy of defect detection. First, image enhancement is performed on the original transmission component image based on two algorithms, Retinex and CLAHE, to enhance the image details while improving the local contrast; then, the image processed by the above two methods is fused to retain the important information of the image and improve the information content of the image; finally, the fused image is input into a convolutional neural network, and the defect detection is carried out after the fully connected layer and softmax classifier. Defect detection. Taking the defective images of transmission line insulators as the experimental dataset, the experimental results show that the method proposed in this paper can effectively improve the accuracy of fault diagnosis of transmission components.