Multispectral polarimetric separability of urban materials

Abstract

Polarimetric imagers (Polaris Sensor Technologies) mounted on an airplane acquired remotely sensed field data in the visible (Vis), shortwave infrared (SWIR), and longwave infrared (LWIR) bands. These airborne data focused on manmade urban materials in settings that also included natural materials such as vegetation. Initial analyses indicated that the Vis degree of linear polarization provided the greatest success in distinguishing between natural and man-made materials, that materials that differed more widely in inferred composition exhibited larger Michelson contrast, and that the Stokes parameters S1 and S2 aided separability among urban materials. The same imagers mounted on a 2 m diameter goniometer then acquired laboratory data of selected urban materials representative of the remotely sensed materials, and additional instruments characterized the compositions of the materials. Preliminary analyses of these laboratory measurements improved the statistics of multispectral polarimetric separability and exhibited dependence upon composition, while confirming and extending the field separability results.