Enhancing night time videos using the non-subsampled contourlet transform

摘要

Video contrast enhancement is one of the most important and difficult components in video processing. In order to improve the visual appearance of the night time video and to provide a "better" transform representation for automated video processing, we present a new multi-scale method for night time video contrast enhancement based on the non-subsampled contourlet transform (NSCT). We modify the contourlet coefficients in corresponding subbands via a nonlinear function and take the noise into account for more precise reconstruction and better visualization. Experimental results show that the enhanced night time video by the proposed method improves visual quality and also is better than other conventional techniques.