Analysis of laplacian distribution model for virtual channel in distributed video coding

Chun Yuan,Xiaoguang You,Sheng Fang
DOI: https://doi.org/10.1109/CECNet.2012.6202064
2012-01-01
Abstract:Distributed source coding theory has long promised a new method of encoding video that is much lower in complexity than conventional video methods. In the distributed framework, the decoder is usually tasked with modeling the correlation noise and exploiting the statistics. In this paper, we first analyze the suitability of laplacian distribution in correlation noise model (CNM), which is the key part of DVC, and then we propose a novel distribution model for correlation noise. The result shows that laplacian distribution model couldn't reflect motion intensity very well and our proposed model performs better in Rate Distortion (RD).
What problem does this paper attempt to address?