An effective underwater image enhancement based on the fusion of CLAHE and image constancy sharpening technique

The non-uniform illumination in underwater causes degradation of images as when light propagates through water, it gets scattered and absorbed due to various factors. The captured underwater images are mainly suffer from color distortion, color cast, low contrast, poor illumination, and blurred details due to the optical properties of water. Underwater image enhancement plays a preprocessing stage for high-level vision tasks. Therefore, to restore the degraded image and for visual enhancement, we proposed a fusion method for an effective output. This method generates two versions of color-corrected input image and one is given for contrast limited adaptive brightness preservance histogram equalization(CLABPHE) and the other for uniform illuminated sharpened image. The two are fused using their weight maps: Luminance weight map, and Saliency detection weight map, and they are normalized and then multi-layered fused using Laplacian pyramid-based decomposition to get an enhanced output image. The quantitative and qualitative evaluation of the proposed method works effectively for several datasets. The qualitative evaluation shows our method is better than the other algorithms compared.