Active Polarimetric Imaging for Estimation of Scene Geometry

Active imaging polarimetry for remote sensing applications has received significant attention recently. Such systems use a variably-polarized active light source to illuminate target objects. Multiple images with different polarizations are then captured and used to build Stokes vectors which are, in turn, used to estimate the refraction indices of materials as well as the relative geometry of the target object. The applications facilitated by active polarimetry include target detection, object recognition, shape extraction and material classification. Unfortunately, this estimation problem requires us to find the solution to a system of nonlinear equations using an iterative optimization technique. In this paper, we introduce a methodology for finding and validating good solutions to this optimization.