Application of the PReMA algorithm to remove the polarization-albedo bias from remote sensing imagery

We applied the recently developed PReMA (Polarization Relative to the Maximum Attainable) algorithm^{1, 2} to test its ability to remove the polarization-albedo bias from remote polarimetric imagery. Our results show that, even though PReMA was derived for 45° specular geometry, it is effective at non-specular geometries as well. Application of the PReMA transform to image data removed the polarization-albedo bias exhibited by the standard DoLP (Degree of Linear Polarization polarization) metric, and it also added a significant boost to the polarimetric signature of all targets, especially the brighter ones. The PReMA algorithm should thus improve the detectability of polarized targets in remote sensing data. We demonstrate that application of the PReMA transform results in pixel values that are proportional to surface roughness, allowing analysts to perform improved surface characterization on remote polarimetric data.