Abstract

The environmental damage and global change caused by wildfires are fast becoming issues of local, regional and global concern. Remote sensing techniques are an ideal tool to use for monitoring these threats. Unfortunately, most existing satellites were built for other purposes and are less than ideal for the problems associated with detecting, assessing and monitoring fires. This thesis examines the phenomenology of fires and rank orders several non-traditional detection strategies that have the promise of being more cost effective and less susceptible to false alarms than existing satellite systems and algorithms. The possibility of detecting fires with multi-band "color" measurements is recommended as an alternative to single-band threshold methods.

Library of Congress Subject Headings

Wildfires--Remote sensing; Forest fires--Remote sensing; Remote sensing; Algorithms; Image processing

Publication Date

1-1-2002

Document Type

Dissertation

Student Type

Graduate

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Schott, John

Advisor/Committee Member

Kremens, Robert

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: SD421.375 .F673 2002

Campus

RIT – Main Campus

Download

Share

COinS