Insulator infrared image denoising using Gaussian Mixture Model with adaptive component selection

Infrared technology has been applied widely to monitor the high voltage insulator in electric power system. However, the insulator infrared image is always contaminated by noise. In this paper, an effective denoising algorithm for contaminated insulator infrared images is proposed. First, the component-wise expectation maximization is used to adaptively select the optimal number of Gaussian mixture model (GMM) components, and a more accurate model is obtained. Then an insulator infrared image denoising algorithm based on maximum a posteriori (MAP) estimation is derived. Finally, the validity of the proposed algorithm is tested. Experimental results we obtained confirm the superiority of the proposed algorithm over the traditional EM-based GMM methods and threshold-based denoising methods.