LAUSR

Side match vector quantization (SMVQ) is an effective coding technique and it has been widely used in low bit rate image compression and data hiding techniques. It utilizes the correlations between neighboring blocks to better generate a small fixed-size state codebook for each input vector. However, compared with vector quantization (VQ) scheme, the quality of SMVQ-coded image is substantially decreased when bit rate (BR) becomes low. This paper proposes a new side match vector quantization scheme which is based on using extend state codebook (ESCSMVQ). Experimental results show that comparing to conventional SMVQ, the bit rate of ESCSMVQ can be significantly reduced meanwhile keeping the exactly same quality of VQ-coded image.