Global optimal reconstruction of underwater polarization imaging using low-pass frequency-filtering constraint

Polarization imaging is an effective approach for clear imaging in underwater scattering environments. However, state-of-the-art methods mostly assume that the degree of polarization (DOP) of backscattered light is uniform, and require initial value optimization or average calculation of self-selecting background regions. Consequently, these not only compromise the imaging quality, but also introduce an additional step of human-computer interaction that hinders its real-world application. In this paper, a global optimal reconstruction method for underwater polarization imaging is proposed using low-pass frequency-filtering constraint. This method estimates the DOP globally with a two-dimensional matrix, and identifies the maximum and minimum intensities of backscattered light under the frequency domain constraint. Compared to cutting-edge polarization imaging methods, our method distinctly provides more accurate backscattered information, and alleviates the complex process of self-selecting the background region, thus resulting in significantly improvement of imaging quality. Qualitative and quantitative experimental results show that the proposed method can enhance the contrast of the image while preserving image details, and it is not only applicable to underwater targets of different materials, but also effective in underwater media of different turbidities.