Indian Journal of Science and Technology

Abstract

It has been observed that characteristics of underwater images are similar to day hazy images under poor lighting situation. So the problem of image visibility and color cast for underwater scenes is still challenging and demands more research attempts in this direction. This paper proposes an approach for shallow underwater visibility improvement which utilizes a combination of histogram equalization, gamma correction, and dark channel prior with morphological operation, in order to achieve effective haze subtraction and avoid halo effects in a single image with a complex structure. Our methodology, also considerthe effect of an adaptive gamma correction technique for further refinement of transmission map. The outcomes shows better peak signal to noise ratio and contrast noise ratio values over existing state-of-the-art methods based ondark channel prior. Results shows that our proposed approachin turn can be utilized in various underwater applications such as forensic image recognition, telecommunication cables recognition etc.
Keywords: Gamma, Haze, Histogram, Morphological, Underwater, Visibility