Wyner-Ziv Video Coding for Low Bitrate Using Spiht Algorithm

Abstract

Distributed video coding (DVC) is a new compression method based on two key information theory results: Slepian-Wolf and Wyner-Ziv theorems. In this paper, we proposed a Wyner-Ziv video coding scheme based on wavelet transform and set-partition in hierarchical trees (SPIHT) which can exploit the spatial, temporal and statistical correlations of the frame sequence. In our scheme, we use Discrete Wavelet Transform (DWT) before quantization, then only coefficients of low frequency subband are Wyner-Ziv encoded using turbo codes, and all coefficients of high frequency subbands in these frames are coded by the SPIHT algorithm. At the decoder, side-information generated through interpolation was used to conditionally decode the Wyner-Ziv frames. Obtained results show that proposed scheme performs better than intra coding scheme only used SPIHT algorithm especially in terms of decoding efficiency at a correspondingly low bit rate.