Optimal contrast enhancement in long distance snapshot polarimetric imaging through fog

Optimally enhanced vision of a polarized lightmark in obscured weather conditions (fog, haze, cloud) is reported when imaged over long distances (above 1 km) using a snapshot polarimetric camera. We derive and experimentally validate an optimal adaptive polarimetric representation, whose expression is shown to depend on the correlation of the noise fluctuations in the two orthogonal polarimetric images. We quantitatively compare the gain (experimental and theoretical) in contrast with respect to standard intensity imaging, and standard polarimetric representations. Lastly, we discuss efficient implementation strategies for automated detection in real-time in obscured weather conditions.