A robust color enhancement for underwater images

Abstract

Due to various complex factors in underwater environments. Several issues must to be considered in obtaining clear underwater images, including color cast, low contrast, blurry details, haze, climate, environment and human factors can significantly affect the analysis and research applications of underwater images. The primary causes are the atomization effect caused by dispersion and the color cast caused by inconsistent energy attenuation of each wavelength as light propagates through water. Different original underwater images exhibit varying characteristics, such as diverse color distributions or illumination levels. Applying a single enhancement method to all underwater images may lead to inconsistent results, where some images are effectively enhanced while others are not. To address this issue, this research proposed a novel 2 stage classification-based mechanism after images were dehazed. We employ four image enhancement techniques with proper conditions and proposes a classification-based mechanism for enhancing underwater images. Furthermore, a color correction method based on an improved Gray-World white balance algorithm is introduced. We address the limited dynamic range and brightness issues in underwater images. Quantitative and qualitative results show that our research performs relatively well in the Underwater Image Enhancement Benchmark (UIEB) dataset compared to other recent methods. It is expected to be applied to different types of underwater work and environments, reducing the severe degradation issues commonly encountered in underwater images.