Modification of Lab color model for minimizing blue-green illumination of underwater vision system

Deep underwater images suffer from low contrast and blue-green illumination which leads towards restriction to the visibility of the objects. Most of the previous proposed enhancement techniques for underwater vision system applications improve the image contrast but blue-green illumination retains in the images. This paper discusses the improvement of underwater image contrast with concentration in modification of Lab color model. The modification of Lab color element in this color model based on shifting the entire pixels of the image to another shifting values, in which, the new shifted values exhibit an improvement in terms of image contrast and minimizing blue-green illumination based on optimal information in the image. The proposed method integrates two main steps, namely dark-stretched image fusion (DSF) and pixel distribution shifting (PDS). In DSF step, dark channel image is stretched, divided into two channels before it is fused together. Next, the image pixel distribution in Lab color model is shifted towards more natural view based on human visual system. Qualitative evaluation indicates a significant improvement of image contrast in output images.