|Abstract:Infrared thermal image region of interest(ROI) extraction has important significance for fault detection, target tracking and so on. In order to solve the problems of many infrared thermal image disturbances, artificial markers and low accuracy, a ROI of infrared thermal image extraction algorithm based on fusion of multi-modal feature map was proposed. Multi-modal feature maps were constructed by contrast, entropy, and gradient features, and region filling was performed to achieve ROI extraction. Apply the new algorithm to the actual collected photovoltaic solar panel image. The results show that the new algorithm has the advantages of high average precision (93.0553%), high average recall (90.2841%), F index and J index are better than Grab Cut, less artificial marks, etc.. It can be effectively used for ROI extraction of actual infrared thermal images.