MULTIPLEDESCRIPTIONSHIFTEDLATTICEVECTOR QUANTIZATIONFOR PROGRESSIVEWAVELET IMAGE CODING

Huihui Bai

Abstract

Multiple description (MD)coding isapromising alternative forrobust transmission ofinformation overnon-prioritized and unpredictable networks. Furthermore, practical variable-bandwidth channels alsorequire finegrain scalability ofthedescriptions (bit streams). Inthis paper, according tothegeometrical structure andthespecial relationship oflattice vector quantizers, aMD quantizer called shifted lattice vectorquantization (SLVQ)is employed inMD imagecoding torealize progressive transmission overunreliable channels. Inviewofthe characteristics ofwavelet coefficients indifferent frequency subbands, besides anappropriate construction ofwavelet coefficient vectors, thealgorithm ofmodified zerotree coding isalso applied toimprove compression performance. Experimental results validate theeffectiveness ofthe proposed schemewithbetter performance thantheother schemes basedonMD scalar quantization forprogressive transmission. IndexTerms-lattice vectorquantization, multiple description coding, image coding, wavelet transform
BibTeX
@inproceedings{bai2006multipledescriptionshiftedlatticevector,
  author = {Huihui Bai},
  title = {{MULTIPLEDESCRIPTIONSHIFTEDLATTICEVECTOR QUANTIZATIONFOR PROGRESSIVEWAVELET IMAGE CODING}},
  booktitle = {ICIP},
  year = {2006},
  doi = {10.1109/icip.2006.312522},
}

Loading related content...

Figure