A fast search algorithm for mean-removed vector quantization using edge and texture strengths of a vector

Authors:
Jim Z. C. Lai;Yi-Ching Liaw
Affiliations:
Department of Computer Science, National Taiwan Ocean University, Keelung 202, Taiwan, ROC;Department of Computer Science and Information Engineering, Nanhua University, Chiayi 622, Taiwan, ROC
Venue:
Image and Vision Computing
Year:
2005

Citing 9
Cited 0

Vector quantization and signal compression

Vector quantization and signal compression
Artifact reduction of JPEG coded images using mean-removed classified vector quantization

Signal Processing
Finite-state vector quantization for waveform coding

IEEE Transactions on Information Theory
Lossless compression of VQ index with search-order coding

IEEE Transactions on Image Processing
Inverse error-diffusion using classified vector quantization

IEEE Transactions on Image Processing
An efficient encoding algorithm for vector quantization based on subvector technique

IEEE Transactions on Image Processing
Fast-searching algorithm for vector quantization using projection and triangular inequality

IEEE Transactions on Image Processing
Side match and overlap match vector quantizers for images

IEEE Transactions on Image Processing
Predictive classified vector quantization

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, a fast search algorithm for mean-removed vector quantization is proposed. Two inequalities are used to reduce distortion computations. Our algorithm makes use of a mean-removed vector's features (edge and texture strengths) to reject many unlikely codewords and it has the same image quality as the full search method. Experimental results show that our algorithm is much better than the full search method in terms of computing time and the number of distortion calculations. Comparing with the full search method, our method can effectively reduce the computing time by 60.2-94.2% and the number of distortion computations by 78.6-97.9% for the codebook sizes of 64-2048. As far as we know, our method is the first of its kind to reduce the encoding time for mean-removed vector quantization.