A New Approach to Video Coding Based on Discrete Wavelet Coding and Motion Compensation

Discrete Wavelet Transform (DWT) is a recent and powerful mathematical tool which has stimulated many developments in several scientific and technical fields, particularly in signal and image processing. Its multiresolution properties are a major advantage. However, in video coding, motion management using DWT is a major challenge due to the DWT decomposition algorithms which does not preserve translation invariance. This means that, a translation of the original frame does not necessarily imply a translation of the corresponding wavelet coefficients. This is one of the reasons which justify, until now, the absence of the DWT-based video compression standard. However, in Discrete Cosine Transform based video compression standard such as MPEG.2, H.263 and H.264 the motion management in video sequences has reached the maturity. This paper introduces a new framework for color video coding that which can reduce the cost of storage and the bandwidth transmission of video file. The new method is a sub-band coding approach that employs Discrete Wavelet Transform (DWT) and is based on a separate sign coding (SSC) from wavelet coefficients amplitude. Furthermore, it uses motion compensation technique (MC) to overcome the problem caused by the lack of invariance translation DWT which leads to false motion vectors. The video codec that we propose in this work is then called Separate Sign Coding with Motion Compensation (SSCMC). An assessment of the proposed scheme was operated on a set of standard video sequences and the results obtained indicate that SSCMC outperforms H.264-AVC in terms of PSNR quality assessment and in terms of perceptual metric as structural similarity metrics (SSIM). Furthermore, the decoded video sequences visualized presents a competitive visual quality compared to H.264-AVC standard.