Abstract

This dissertation develops a polarimetric thermal infrared (IR) framework within the Digital Image and Remote Sensing Image Generation (DIRSIG) software tool enabling users in the remote sensing community to conduct system level trades and phenomenology studies. To support polarized reflection and emission modeling within DIRSIG, a generalized bi-directional reflectance distribution function (BRDF) is presented. This generalized form is a 4x4 element Mueller matrix that may be configured to resemble the commonly utilized Beard-Maxwell or Priest-Germer BRDF models. A polarized emissivity model is derived that leverages a hemispherical integration of the polarized BRDF and Kirchoff's Law. A portable experimental technique for measuring polarized long-wave IR emissivity is described. Experimental results for sixteen target and background materials are fit to the polarized emissivity model. The resulting model fit parameters are ingested by DIRSIG to simulate polarized long-wave infrared scene phenomenology. Thermally emitted radiance typically has a vertical polarization orientation, while reflected background radiance is polarized horizontally. The balance between these components dictates what polarized signature (if any) is detected for a given target. In general, specular targets have a stronger emission polarization signature compared to diffusely scattering targets consistent with visible polarimetry findings. However, the influence of reflected background radiance can reduce the polarimetric signature of specular targets below a detectable threshold. In these situations, a diffusely scattering target may actually exhibit a polarization signature stronger than a specular target material. This interesting phenomenology is confirmed by experimental scene collections and DIRSIG simulations. Understanding polarimetric IR phenomenology with this level of detail is not only key for system design, but also for determining optimal collection geometries for specific tactical missions.

Library of Congress Subject Headings

Infrared imaging--Computer simulation; Remote sensing--Computer simulation; Polarimetric remote sensing; Polarimetry

Publication Date

5-2007

Document Type

Dissertation

Student Type

Graduate

Degree Name

Imaging Science (Ph.D.)

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

John R. Schott

Advisor/Committee Member

Carl Salvaggio

Advisor/Committee Member

Anthony Vodacek

Comments

Physical copy available from RIT's Wallace Library at TA1570 .G37 2007

Campus

RIT – Main Campus

Share

COinS