Power Inspection Robot Based on Equipment Fault Spectrum Image Recognition

Abstract

The power industry is the guarantee of people's basic life, so the security of which is crucial. Transmission line is one of the important parts of power grid. Line inspection can ensure the safe operation of transmission line, find the hidden danger of power line in time, prevent the occurrence of power line fault, or narrow the scope of fault. Due to the equipment of infrared thermal imager on inspection robot, the infrared image of power facilities contains complex background information, including land, buildings and other interference areas. In this paper, a high-precision segmentation algorithm based on local gray features of spectral region is proposed. Firstly, the infrared image of spectrum is preprocessed. Then, the image is dried based on Gaussian model. Finally, the algorithm based on local standard deviation feature, morphological expansion and size and location information filtering algorithm are used to achieve accurate spectral region segmentation.