Abstract: AbstractIn this paper, a dynamic state codebook-based side-match vector quantization (DSMVQ) is presented. The super codebook generation procedure is extended to explore the characteristics of codewords and to determine the dynamic state codebook sizes. In encoding, the 'near' codeword from the super codebook is selected based on the side-match distortion measurement. The size of the state codebook is retrieved according to the 'near' codeword. Then, side-match prediction encoding is employed to find the appropriate codeword from the state codebook. The experimental results show that the proposed scheme provides a lower encoding bit rate than classified SMVQ (CSMVQ) and two-level CSMVQ, and maintains better image quality than SMVQ and three-sided SMVQ. Furthermore, the adaptive encoded image quality is obtained according to the predefined VQ approximation rate. The adaptive encoding provides a possible solution to the trade-off between the encoding bit rate and the image quality.Keywords: NEIGHBOURING SIMILARITYSIDE-MATCH VECTOR QUANTIZATIONVECTOR QUANTIZATION
Publication Year: 2004
Publication Date: 2004-06-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot