Statistical estimation problems of atmospheric correction techniques for thermal infrared imagery have been studied. A revised multiple-view- angle atmospheric correction technique has been developed and tested using the LOWTRAN radiative transfer model as truth. Its average absolute temperature prediction accuracy for an independent data set is 0.8 K for long-wave infrared imagery and 1.0 K for mid -wave infrared imagery, when error-free data are assumed. The benefit of robust, resistant regression estimators was studied using Monte Carlo simulations having real -world measurement errors and data outliers. An error propagation analysis showed that 1.0-3.3 K of rms error is likely given reasonable data set sizes and robust estimators.

Library of Congress Subject Headings

Infrared imaging; Thermography; Remote sensing

Publication Date


Document Type


Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)


Schott, John

Advisor/Committee Member

Fairchild, Mark

Advisor/Committee Member

Salvaggio, Carl


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: TA1570 .M47 1992


RIT – Main Campus