Camouflaged target detection using real-time video fusion algorithm based on multi-scale transforms

In this paper, we propose a novel video fusion algorithm for real-time detection of camouflaged targets. Initially, the targets are detected by using a novel target detection method by applying conventional image thresholding methods in the wavelet domain. The thermal information corresponding to the detected targets are transferred to the output fused image along with the background information by using novel region-based fusion rules. The dominance of the proposed algorithm in terms of the detection of the camouflaged targets is visibly evident from the quantitative and qualitative results of the output fused video.