Underwater Image Enhancement Based on Optimal Contrast and Attenuation Difference

Abstract

The quality of underwater images is often marred by noticeable color casts and blurring, resulting from complex phenomena such as differential light attenuation based on wavelength, veiling light scattering, and light scattering by plankton and suspended particles in water. In this paper, we propose an effective underwater image enhancement method called Optimal Contrast and Attenuation Difference (OCAD) to tackle these issues. The OCAD method employs a two-step approach. First, it estimates a coarse transmission map by optimizing contrast and minimizing information loss during image mapping. This transmission map is then refined using a combination of dark channel prior and guided filter refining techniques to improve its accuracy. Second, OCAD estimates the veiling light by considering the differential attenuation of red, green, and blue light underwater. This estimated veiling light is utilized to partially mitigate the color cast caused by the attenuation effect, while also addressing the blurring based on an imaging model. To further enhance the image quality, we introduce the Gray World approach to correct the color and obtain a deblurred and color-corrected underwater image. To evaluate the performance of our proposed method, extensive experiments are conducted on well-established underwater image datasets, including the EUVP, UIEB, and RUIE datasets. Comparative analysis against state-of-the-art and classical underwater image enhancement methods reveals that our OCAD algorithm significantly enhances underwater images and outperforms other methods in terms of image quality improvement.