Research And Application Of Infrared-Image Matching Technology In The Power Equipment Inspection System

The power equipments are main equipment for the factory, so they are very important. The power equipments are always works in the high voltage, strong current and high temperature environment, in the meantime, they also suffer the sunshine, rain and even work in the worse environment in the long time. So it is very easy that the power equipments have some problems. To make sure the safety and reliable power generation, we must carry out the power equipment inspection. The traditional inspection approach is inefficient, so it has urgent need for the more advanced and matured inspection method. This thesis mainly discuss the power equipment inspection system based on the infrared image matching technology, and it will analyze from two aspects, the first one is the architecture of power equipment inspection, the second one is infrared image matching algorithms.Infrared image matching algorithms will be discussed mainly in this thesis. Because the image acquisition equipment is hardware equipment, it is very easy that the stop position have some little error which render the initially position is different with the position every time the equipment stopped. The system must carry on the image matching algorithms. This thesis analyzes the algorithm in detail which consists of five parts. The first part is image segmentation algorithms; the second part is morphology skeleton extraction algorithms; the third part is feature point extraction algorithms; the fourth part is feature point matching algorithms; and the final part is to find the coordinate relation between the reference image and the matching image. The experiment show that this algorithms have high accuracy, but which still needed to be improved. The feature point matching algorithms should be improved and discussed in further.