Masters Theses

Date of Award

12-1993

Degree Type

Thesis

Degree Name

Master of Science

Major

Computer Science

Major Professor

Bruce Whitehead

Committee Members

Dinesh Mehta, Al Pujol

Abstract

Fractal image compression describes an image by storing the relationship between different regions of an image in the form of contractive affine transforms. This technique provides extremely high compression rates while maintaining good image fidelity and rapid access to the image. The disadvantage of this technique is the intensive computational requirements for image compression. Computational requirements can be reduced by minimizing the number of comparisons that must be made.

A set of domains and ranges were defined over each image compressed. For each range a domain pool was defined consisting of all domains within a given distance (in pixels) from the range. As the size of the domain pool was varied compression ratios and image qualities were measured.

This research concluded that, based on the compression of three images, a distance of approximately 60 pixels resulted in an acceptable compromise between image fidelity and computational requirements. At distances less than 60 pixels increasing the distance results in large benefits in terms of compression ratio and image fidelity. There was little benefit from using distances greater than 60 pixels.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS