Three-Description Image Coding Using Optimal Dead-Zone Lattice Vector Quantization

Multiple description coding is an effective way to transmit images over unreliable network or multiple channels. In this paper, a three-description image coding scheme is proposed, involving an optimal index assignment to obtain a mapping table that can reduce side distortion and dead-zone lattice vector quantization to quantize different coefficients more efficiently. The aim is to provide more robustness and present a good reconstructed image quality with bit rate as low as possible for unreliable network. The simulation results of standard image show that with consideration of wavelet coefficients characteristic the proposed scheme can reduce side distortion and reduce expected distortion of the system at high loss rate and compression efficiency.

[1]  N. J. A. Sloane,et al.  Multiple-description vector quantization with lattice codebooks: Design and analysis , 2001, IEEE Trans. Inf. Theory.

[2]  Xiaolin Wu,et al.  Optimal Design of Multiple Description Lattice Vector Quantizers , 2006, ArXiv.

[3]  Vinay A. Vaishampayan,et al.  Design of multiple description scalar quantizers , 1993, IEEE Trans. Inf. Theory.

[4]  Xiaolin Wu,et al.  Index assignment design for three-description lattice vector quantization , 2006, 2006 IEEE International Symposium on Circuits and Systems.

[5]  Yao Zhao,et al.  Optimized multiple description image coding using lattice vector quantization , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[6]  N. J. A. Sloane,et al.  Multiple description lattice vector quantization , 1999, Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096).