Date of Award

9-14-2017

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Department of Engineering Physics

First Advisor

Kevin C. Gross, PhD.

Abstract

This research introduces two novel oblique longwave infrared atmospheric compensation techniques for hyperspectral imagery, Oblique In-Scene Atmospheric Compensation (OISAC) and Radiance Detrending (RD). Current atmospheric compensation algorithms have been developed for nadir-viewing geometries which assume that every pixel in the scene is affected by the atmosphere in nearly the same manner. However, this assumption is violated in oblique imaging conditions where the transmission and path radiance vary continuously as a function of object-sensor range, negatively impacting current algorithms in their ability to compensate for the atmosphere. The techniques presented here leverage the changing viewing conditions to improve rather than hinder atmospheric compensation performance. Initial analyses of both synthetic and measured hyperspectral images suggest improved performance in oblique viewing conditions compared to standard techniques. OISAC is an extension of ISAC, a technique that has been used extensively for LWIR AC applications for over 15 years, that has been developed to incorporate the range-dependence of atmospheric transmission and path radiance in identification of the atmospheric state. Similar to ISAC, OISAC requires the existence of near blackbody-like materials over the 11.73 micrometer water band. RD is a newer technique which features unsupervised classification of materials and identifies the atmospheric state which best detrends the observed radiance across all classes of materials, including those of low emissivity.

AFIT Designator

AFIT-ENP-DS-17-S-030

DTIC Accession Number

AD1051608

Download

Share

COinS