An automatic diagnostic method of abnormal heat defect in transmission lines based on infrared video

Infrared videos from transmission line inspection of UAV have a large amount of data with low SNR (Signal to Noise Ratio). Identifying heat defects automatically using infrared videos is difficult and inefficient. In this paper, an advanced method is proposed. Firstly, points with excessive value of the component regions according to conductors are analyzed in their neighbors to obtain defect points. Then, the heat region of each defect points is segmented, and defect type of which is distinguished automatically by target accounting, skeleton, convex defects, position of lead wire, LBP feature vector. To solve the problem of low efficiency, an infrared video is divided into segments encompassing tower (SETs) and segments don't encompassing tower (SNETs). Heat defects of clamps, lead wire joints, insulators are processed using SETs. Experiment shows that defect locating accuracy is 91.4%, false alarm rate is 12.3%. Classification accuracy of the located defects is 82.3%. Then, this method is effectiveness and robustness.