DocumentCode :
352386
Title :
Rate-distortion adaptive vector quantization for wavelet image coding
Author :
Gu, Qun ; Budge, Scott E.
Author_Institution :
Dept. of Electr. & Comput. Eng., Utah State Univ., Logan, UT, USA
Volume :
6
fYear :
2000
fDate :
2000
Firstpage :
1903
Abstract :
We propose a wavelet image coding scheme using rate-distortion adaptive tree-structured residual vector quantization. Wavelet transform coefficient coding is based on the pyramid hierarchy (zero-tree), but rather than determining the zero-tree relation from the coarsest subband to the finest by hard thresholding, the prediction in our scheme is achieved by rate-distortion optimization with adaptive vector quantization on the wavelet coefficients from the finest subband to the coarsest. The proposed method involves only integer operations and can be implemented with very low computational complexity. The preliminary experiments have shown some encouraging results: a PSNR of 30.93 dB is obtained at 0.174 bpp on the test image LENA (512×512)
Keywords :
computational complexity; discrete wavelet transforms; image coding; optimisation; rate distortion theory; transform coding; tree data structures; vector quantisation; LENA test image; PSNR; adaptive vector quantization; coarsest subband; finest subband; integer operations; pyramid hierarchy; rate-distortion optimization; tree-structured residual vector quantization; very low computational complexity; wavelet image coding; wavelet transform coefficient coding; zero-tree relation; Discrete transforms; Image coding; Internet; Multimedia communication; PSNR; Rate-distortion; Shape; Testing; Vector quantization; Wavelet coefficients;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech, and Signal Processing, 2000. ICASSP '00. Proceedings. 2000 IEEE International Conference on
Conference_Location :
Istanbul
ISSN :
1520-6149
Print_ISBN :
0-7803-6293-4
Type :
conf
DOI :
10.1109/ICASSP.2000.859200
Filename :
859200
Link To Document :
بازگشت