Underwater Image Enhancement via Color Balance and Stationary Wavelet Based Fusion

Abstract

Underwater imaging is one of the most interesting and popular field for researchers and scientists now a day. Generally, underwater images are suffered from poor visibility and color shift because of absorption of different wavelength lights and scattering of water droplets. To address these problems, this paper proposed a novel framework based on fusion principle for enhancing of underwater images. Our strategy consists of four main steps. In first step, we perform color correction using white balancing based on gray world algorithm to remove the unwanted color cast. Then we processed the white balanced image using contrast enhancement techniques i.e. gamma correction of optimal value is 1.5 and sharpening. After second step, both versions of images of contrast enhancement techniques are processed through weight mapping of three categories i.e. luminance, chromatic and saliency. Finally, both versions of images are fused through stationary wavelet based fusion rule to get better quality final enhanced image. Our quantitative and qualitative analysis reveals that our approach outperforms better as compared to state-of-the-art algorithms. Our corroboration also demonstrates that the enhanced images are characterized by improved edge sharpness, global contrast and clear visible dark regions.