Subband VPIC with classified joint vector quantization

Author

Chan, K.W. ; Chan, K.L.

Author_Institution

Dept. of Electron. Eng., City Univ. of Hong Kong, Kowloon, Hong Kong

Volume

3

fYear

1996

fDate

16-19 Sep 1996

Firstpage

383

Abstract

Subband coding (SBC) was combined with visual pattern image coding (VPIC) to eliminate the blockiness and reduce the jagged edge artifact. A new set of 2×2 patterns was designed to code the highly compacted subband signals. The MSEs are significantly reduced by coding only the lowest frequency band. By making use of the inter-band correlation among various subband signals, a classified joint vector quantization (CJVQ) scheme was derived for coding the high frequency and residue subbands, using the pattern index in the low frequency band (LFB) as the classifier. The PSNR is further enhanced by up to 1.7 dB with less than a 8% decrease in the compression ratio

Keywords

coding errors; correlation methods; decoding; image classification; image coding; vector quantisation; visual perception; MSE; PSNR; classified joint vector quantization; compression ratio; decoding; edge artifact reduction; high frequency subbands; interband correlation; low frequency band; pattern index; residue subbands; subband VPIC; subband signals; Arithmetic; Decoding; Distortion measurement; Filters; Frequency; Hardware; Image coding; Signal analysis; Signal synthesis; Vector quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing, 1996. Proceedings., International Conference on

Conference_Location

Lausanne

Print_ISBN

0-7803-3259-8

Type

conf

DOI

10.1109/ICIP.1996.560511

Filename

560511