Thermal image segmentation by K -means clustering algorithm

Thermal images of panels are captured using a FLIR T420bx thermal imager. The thermal images of panels are analyzed by segmenting the image

The HSV, CMY, NTSC, YCbCr color spaces are not well suited for describing colors in terms that one practical for human interpretation. So, the RGB thermal image has to be converted into HIS image for better visibility to indentify the defects presented in the panel. and apply the K -means clustering algorithm to segment the image and diagnose the defects in the PV panel

Reference:
Uma, J., Muniraj, C., and Sathya, N., "Diagnosis of Photovoltaic (PV) Panel Defects Based on Testing and Evaluation of Thermal Image," Journal of Testing and Evaluation, https://doi.org/10.1520/JTE20170653. ISSN 0090-3973.