Abstract

This study examined the use of textural analysis to assist in the segmentation of synthetic aperture radar (SAR) imagery. A previous study, which dealt with imagery collected passively by electro-optical systems in the visible and infrared portions of the electromagnetic spectrum, tested a method of identifying optimal subsets of texture features from a large pool of possibilities. Three new features were added to this pool and tested against the others to determine their specific utility in the segmentation of SAR imagery. Their value is demonstrated by consistent appearance in optimal subsets of texture features, and by their contribution of up to 8% to image classification accuracy.

Library of Congress Subject Headings

Synthetic aperture radar; Image processing--Digital techniques; Filters (Mathematics)

Publication Date

4-1-1994

Document Type

Thesis

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Schott, John

Advisor/Committee Member

Easton, Roger

Advisor/Committee Member

Rhody, Harvey

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: TK6592.S95P345 1994

Campus

RIT – Main Campus

Download

Share

COinS