DocumentCode :
1734306
Title :
Hierarchical multi-stage classified vector quantization
Author :
Chan, K.W. ; Chan, K.L.
Author_Institution :
Dept. of Electron. Eng., City Univ. of Hong Kong, Hong Kong
Volume :
1
fYear :
1995
Firstpage :
468
Abstract :
Two advanced vector quantization schemes were proposed. These schemes made use of the spatial correlation of different block size in different hierarchical levels. In the top level, the image is partitioned into large block size and coded by a small codebook and every index of the codeword acts as the classification index of the succeeding level. In the next level, the differential image is partitioned into smaller blocks, and coded by a number of subcodebooks. These hierarchical, multi-stage and classified coding methods extend the coverage of codewords, thus improved the visual quality, while without penalty in compression ratio
Keywords :
image classification; image coding; image segmentation; vector quantisation; block size; classification index; codebook; compression ratio; differential image; hierarchical multistage classified vector quantization; spatial correlation; subcodebooks; visual quality; Bit rate; Distortion measurement; Image coding; Image quality; Image reconstruction; Image resolution; Mean square error methods; Nonlinear distortion; Signal resolution; Speech coding; Statistics; Vector quantization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Global Telecommunications Conference, 1995. GLOBECOM '95., IEEE
Print_ISBN :
0-7803-2509-5
Type :
conf
DOI :
10.1109/GLOCOM.1995.501971
Filename :
501971
Link To Document :
بازگشت