IMAGE PROCESSING

UNDERWATER IMAGE COLOR CORRECTION

Water absorbs light similarly to an optical filter but with higher variations and complexities. A liquid medium modifies the spectral power distribution of the transmitted light, such that a strong bluish or greenish color cast dominates underwater images. We proposed a new real-time algorithm based on a Tikhonov-type of optimization model in the CIELAB space to adaptively remove the strong color cast and enhance the underwater image quality. Interestingly, this model provides a novel mathematical interpretation of the complementary adaptation theory in psychology.

[He. 2020 (submitted, arXiv preprint)]