Efficient coding algorithms for vector quantization of images

Pao Yue-kong Library Electronic Theses Database

Efficient coding algorithms for vector quantization of images

 

Author: Liu, Wang-kwong, Alan
Title: Efficient coding algorithms for vector quantization of images
Degree: M.Sc.
Year: 1995
Subject: Image processing -- Digital techniques
Coding theory
Data compression (Telecommunication)
Hong Kong Polytechnic University -- Dissertations
Department: Multi-disciplinary Studies
Pages: 1 v. (various pagings) : ill. ; 30 cm
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b1183439
URI: http://theses.lib.polyu.edu.hk/handle/200/2643
Abstract: In the past decade, vector quantization (VQ) has widely been used in image coding systems for its superior rate-distortion performance than the traditional scalar schemes at low bit-rates. In vector quantization of images, the computation concentrates at the encoder. When the codebook size is M and vector dimension is k, the encoding process will require M multiplications and (2k-1)M/k additions for each pixel if Euclidean distortion is used. Such complexity increases exponentially with the bit-rate and vector dimension k, and thus limits the usage of larger codebook sizes and vector dimension. So, many methods have been proposed to reduce the VQ encoding complexity. However, most of these methods search every codeword in the codebook for the minimum distortion codeword of the input vector. They do not make use of the fact that the image data of two adjacent blocks are correlated and so the minimum distortion codewords are probably in the vicinity of each other. In this dissertation, efficient coding algorithms for vector quantization of images. The proposed algorithms should have simple implementation structure, simple computation, low bit-rate requirement and good image quality. First we propose a simple but efficient algorithm, Improved Minimum Distortion Encoding (IMDE) algorithm, to accelerate the encoding process in a vector quantization scheme when a Minimum Squared Error (MSE) criterion is used. A new distortion measure and a new criterion for determining whether the searched codeword is close enough to the minimum distortion codeword are developed. The codebook is re-ordered according to the new distortion measure such that minimum distortion codeword can be found after a few steps and a number of codewords can be rejected without any computations. Next, another fast encoding method called Predictive Mean Search (PMS) algorithm is proposed for image compression. With this method, the minimum distortion codeword can be found by searching only a portion of the codebook and its relative address to the origin of search is sent instead of the absolute address. Three variants of this method are proposed for transmission of the relative address of the minimum distortion codeword. Simulation results show that the number of calculations can be reduced significantly by using the PMS methods.

Files in this item

Files Size Format
b11834390.pdf 2.998Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

     

Quick Search

Browse

More Information