Full search & tree search vector quantization of images

Pao Yue-kong Library Electronic Theses Database

Full search & tree search vector quantization of images

 

Author: Chan, Sai-kin Toby
Title: Full search & tree search vector quantization of images
Degree: M.Sc.
Year: 1996
Subject: Image processing -- Digital techniques
Data compression (Telecommunication)
Coding theory
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=b1230686
URI: http://theses.lib.polyu.edu.hk/handle/200/2049
Abstract: Vector quantization has been receiving a great amount of attention because excellent rate-distortion performance at low bit rate application and the ease of implementing the decoder using the table lookup procedure. A vector quantizer is a function which maps the continuous data space into a finite set of numbers. The numbers address a table (codebook) of codewords which have been chosen to provide a good approximation to the input data. A clustering procedure, such as the generalized Lloyd algorithm, is usually used for codebook generation. In this procedure, the codebook is (locally) optimised for the particular training set, chosen as the representative of the input images. The main problem of VQ systems is the high computational complexity at the encoding which includes the time consuming codebook generation and the best-fit matching encoding process. Our work is to investigate different methods in improving the computational efficiency of the codebook generation process as well as the encoding matching process. We evaluated three methods to speed up the full search codebook design. They are partial distortion and second two use constraints imposed by ordering the codewords. Finally, we explored different algorithms (i.e. Full Search, Tree Structured VQ, Pruned Tree Structured VQ and Entropy Pruned Tree Structured VQ) for the encoding matching process. The performance of these algorithms are simulated in terms of MSE and complexity. The results show that the TSVQ, PTSVQ and EPTSVQ reduce the complexity significantly compared with the Full Search. Also, the PTSVQ and EPTSVQ can achieve a better image quality than others.

Files in this item

Files Size Format
b12306861.pdf 4.859Mb 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