Generalized Equalization Model for Underwater image enhancement

Abstract

The field of underwater image processing has attracted a significant amount of interest over the course of the last several decades, resulting in significant advancements. An underwater images have been essentially characterised by with there poor exposure and since illumination is exponential rate inactivated as it travels through the water, resulting in scenes that are inadequately contrasted and hazy. This is the case because light is absorbed exponentially more as it travels through the water. Due to the fact that blue light has the shortest wavelength, it is able to go the farthest through the water. As a result, blue colour predominates most prominently in photographs that were taken underwater since blue light is able to penetrate the water the longest. In conclusion, the photos suffer from a narrow range of vision, poor contrast, inconsistent illumination, blurring, colour diminishment (a blue look), and noise. Post-processing, which enhances the image's quality and helps increase the sense of depth in underwater photography, is performed with the goal of enhancing the perception. In this study, we suggest an equalisation model that is more generic for the purpose of picture improvement. Based on our investigation of histogram with contrast adjustment balancing interactions, we first propose a generalised equalisation model that integrates image contrast as well as white balanced into an unified model of convex code of image histogram. The experimental findings, both subjective and objective, indicate that the suggested algorithm has favourable performance in the context of applications including picture enhancement, white balance, and tone correction. The convergence speed of the approach that is being offered is also investigated.