Dualband Infrared Images Target Detection Algorithm Based on Multiscale Morphology

According to the features of mid-wave and long-wave infrared images, they are decomposed into morphology pyramid respectively based on the new multiscale mathematical morphology filters proposed in the paper. Some features such as local maximum gray level and average gradient strength of every image are extracted at each level of morphology pyramid. Fuse dualband infrared images based on fusion rule put forward in the paper, and then reconstruct original image and detect target using contrast threshold segmentation. The experiment results show that dualband infrared images target detection algorithm based on multiscale morphology algorithm is better than use mid-wave or long-wave infrared images detect targets alone