Imaging through fog: polarized light propagation modelling

Abstract

Imaging through nebulous media encountered in nature – like fog – is one of the up-to-date research questions in the field of navigation. Among different approaches to solve this problem, the use of polarimetric imaging has been proposed. The aim of this research is to present a polarized light propagation model based on Monte-Carlo simulations and Mie Theory for its use as a tool to study the feasibility and characteristics of polarized light detection techniques in these kinds of media. Studying and obtaining the backscattering Mueller matrix for a turbid environment allows to completely characterize the backscattering characteristics of the media. Thus, it may be used to select the combination of polarization components more useful for each situation. On the other hand, it has been seen that when light first interact with fog, generated backscatter directly blinds the sensor of the imaging device and hides reflected signals of objects being imaged. Using a time-resolved variant of the model, it has been possible to characterize the shape and the influence of the temporal response of a backscattered light pulse, concluding that, apart from the greater light extinction in a more scattering medium, media-backscattering and reflected-object signals are independent