Digital image processing approach using combined wavelet hidden markov model for well-being analysis of insulators

This study employs a digital image processing (DIP) technique for video surveillance (VS) of overhead power distribution line insulators. Such VS technique is one of the promising means for augmenting the process of distribution system automation (DSA). Particularly in semi-urban areas and as well as urban areas in developing countries like India, the overhead distribution lines traverse mostly near the roads and thus VS based on surface vehicular approach equipped with moving cameras seems to be more suitable compared to VS using either remote terminal units (RTUs) or a helicopter. The proposed VS approach has unique advantage due to the fact that a non-technical person can go for patrol using a vehicle so as to capture images of power lines along with insulators for subsequent analysis. In addition, such an approach avoids the difficulties such as blurring of images and camera sight control as in case of aerial VS. Also in case of VS based on RTUs, the maintenance of RTUs becomes a challenging task, particularly because of the outdoor nature of installations. Thus, this study has uniquely employed the use of DIP for video tracking of overhead power lines using template design, feature extraction by applying wavelet transform, and subsequently hidden Markov model is being used for well-being analysis to segregate the damaged insulators from good ones. The case studies validate the efficacy of the proposed methodology for insulator monitoring to augment on-going DSA.