Analysis of spectropolarimetric responses in the visible and infrared for differentiation between similar materials

Abstract

Spectropolarimetric research has focused on target detections of materials that have a high degree of contrast from background materials, such as identification of a manmade object embedded in a vegetative background. This study presents an approach using spectropolarimetric imagery in visible, shortwave infrared, and longwave infrared bands to differentiate between similar natural and manmade materials. The method employs Michelson contrast and Kruskal-Wallis one-way analysis of variance (ANOVA) H-test to determine if a distinction can be found in pairwise comparisons of similar and different materials using the Stokes parameters in the visible, shortwave infrared, and longwave infrared bands. Results showed that similar natural and manmade materials were differentiable in spectropolarimetric imagery using the Michelson contrast and ANOVA. This approach provides a way to use spectropolarimetric imagery to distinguish between materials that are similar to each other.