SBIR/STTR Award attributes
Detection, Recognition and Identification (DRI) of man-made objects in terrestrial and sky backgrounds is often improved with polarization imaging. The polarization signatures of objects depend on several factors including the optical properties, shape and surface characteristics of the scene elements, the sensor look angle, the relative position of illumination sources, and wavelength sensitivity of the imager. Predictive models are needed to help researchers understand the physics behind the polarization signatures of objects, and to provide metrics that predict how DRI algorithms will perform with polarization imagery of various targets, backgrounds and environmental conditions. In this SBIR, Polaris Sensor Technologies proposes to develop a reflective band Polarization Signature Model (PSIM) for the Visible waveband in Phase I, extending to the SWIR band in Phase II. PSIM will be a hybrid empirical/analytical model that predicts polarization signatures of objects based on the modeled/measured incident polarization and the polarization altering properties of the object. The polarization altering properties of the object will be modeled with the Mueller calculus for objects that display diattenuation, retardance and depolarization.
