Autonomous Analysis of Infrared Images for Condition Diagnosis of HV Cable Accessories

Infrared thermography has been used as a key means for the identification of overheating defects in power cable accessories. At present, analysis thermal imaging pictures relies on human visual inspections, which is time-consuming and laborious requires engineering expertise. In order to realize intelligent, autonomous recognition infrared images taken from electrical equipment, previous studies reported preliminary work preprocessing extraction feature parameters, were then train neural networks. However, features required manual selection, reports showed no practical implementations. this contribution, an diagnosis method, based Faster RCNN network Mean-Shift algorithm, proposed. Firstly, trained implement positioning objects be diagnosed images. Then, algorithm image segmentation extract area overheating. Next, parameters determining temperature parts accessories are calculated, made by following relevant condition assessment criteria. Case carried out paper, results show that their regions can located assessed at different camera angles under various background conditions via processing methods proposed paper.