Thermal Defect Detection and Location for Power Equipment based on Improved VGG16

Abstract

Thermal defect affects the normal operation of power equipment, which is crucial to the reliability of the power system. To solve this problem, a thermal defect detection and location method based on neural network is proposed. According to the characteristics of infrared images, a visual geometry group network (VGG16) based on transfer learning is established for temperature recognition. After screening the thermal defect images with abnormal temperature, an improved connected component method is used for defect region location. The results demonstrate that the recognition accuracy of the proposed method is 99.6%. The thermal defect region can be located more accurately. It is significant to realize intelligent detection for power equipment.