An optimized region-based color transfer method for night vision application

Modern night-vision systems like image intensifiers and thermal cameras enable operations at night and in adverse weather conditions. Modern night vision camera provides false-colored fused image as an output which is unnatural in appearance and it is therefore hard to interpret. In this paper, a region-based natural color mapping method for night vision imagery is presented. The proposed method colorizes the night vision imagery by using a combined framework consisting of hill-climbing algorithm for color-based segmentation, non-linear diffusion, region recognition and fuzzy based image fusion techniques. The proposed method is an optimized region-based approach and allows selective color transfer from the natural target color image. The simulation results show that the color fused image obtained by proposed method resembles the natural color appearance and will help the observer by making it more recognizable appearance for better scene interpretation.