Distributed Video Coding (DVC) is a new video coding paradigm based on Slepian-Wolf and Wyner-Ziv theories. The Estimation of the Virtual Channel Model (VCM) greatly influences the coding performance of DVC system. In this work
two algorithms are proposed to dynamically estimate more precise VCM
which have led to better performance of DVC system. For exploiting the temporal dynamic property of VCM
a new frame level VCM estimation algorithm at the decoder is proposed. For each coding frame
the bi-directional vector acquired in side information interpolation is firstly used to get motion compensated version of the key frames. Then the residuals of the key frames and their motion compensated version are used to estimate the VCM of current coding frame. In order to exploit the spatial dynamic property of VCM
a novel block level VCM estimation algorithm is proposed in this paper. The video frames are firstly divided into blocks. Then for each block
the VCM estimation algorithm formerly proposed is employed respectively to get its own VCM. It is proven that the proposed algorithms have gotten more precise VCM. Consequently
they have led to better channel coding performance of DVC
then less parity bits has to be transmitted to decoder
and better coding performance of DVC system are obtained.