Thermal Image Skeleton Extraction and Intelligent Recognition of Power Equipment

The core requirement of power system operation is to ensure that it is safe and stable. In the new era, it has become a trend to use the infrared thermal imager to take pictures of electric equipment and to use the computer to detect thermal faults in thermal images of equipment. In the detection, how to further confirm the fault types through intelligently identifying device types has become a new challenge. A new way of intelligently identifying the device types in the infrared thermal image has been proposed in this paper and a set of processing methods ranging from thermal images to skeleton diagram of electrical appliance have been worked out. Besides, skeleton diagram recognition classification algorithm is preliminarily designed with CNN, which proves the feasibility and supremacy of sketches serving as the characteristic quantities of intelligent identification algorithm.