Abstract

Vector quantization (VQ) has recently emerged as a powerful and efficient technique for digital speech and image coding. The goal of such a process is data compression: to minimize communication channel capacity or digital storage memory requirements while maintaining an acceptable fidelity level of the data. A review of various VQ algorithms and their respective design considerations as applied to color images is given. Fidelity measurements and signal-to-noise ratio calculations are discussed. A modified mean-residual vector quantizer using the LBG design algorithm with color signal preprocessing is described. The algorithm is developed to yield a bit rate of 0.709 bits per pixel per color with the goal of easy implementation even using a simple microcomputer . Photographic and numeric results of original versus compressed-uncompressed color images are presented. Several modifications to the described algorithm are tested with good results .

Library of Congress Subject Headings

Data compression (Computer science); Data compression (Telecommunication); Image processing--Digital techniques

Publication Date

1986

Document Type

Thesis

Student Type

Graduate

Degree Name

Electrical Engineering (MS)

Department, Program, or Center

Electrical Engineering (KGCOE)

Advisor

Joseph DeLorenzo

Campus

RIT – Main Campus

Plan Codes

EEEE-MS

Download

Share

COinS