Insulator Contamination Measurement Based on Infrared Thermal and Visible Image Information Fusion

Abstract

Pollution-induced flashover has a serious threat on the safe and reliable operation of power systems In this paper infrared image methods and visible image methods are utilized comprehensively to find out porcelain Insulator contamination defect in a laboratory This method is better than the traditional visible image and infrared thermal image in the detection of insulator pollution. The surface temperature and color information of insulators are represented by infrared thermal image and visible image respectively. To better preserve the useful information from source images in this paper, a novel image decomposition method based on latent low-rank representation (LatLRR) is developed, which is simple and effective. This modified image fusion method based on the guided image filtering is used, which can effectively extract small-scale texture detail information of the visible image and the large-scale edge information of the infrared image. Firstly, the source images are decomposed into low-rank parts (global structure) and saliency parts (local structure) by LatLRR. Then, the low-rank parts are fused by weighted-average strategy, and the saliency parts are simply fused by sum strategy. Finally, the fused image is obtained by combining the fused low-rank part and the fused saliency part. Experiments demonstrate that the proposed method is better than the traditional multi-scale decomposition method based on image fusion algorithms in both subjective and objective evaluations for insulator contamination measurement and we shall present the results.