Underwater Image Enhancement by Modified Dark Channel Prior

The study of underwater image analysis has generated a lot of interest and demonstrated significant improvements. Images taken underwater suffer from unbalanced colors and low contrast. DCP is one of the algorithms used in image enhancement. In this study, an algorithm, Modified Dark Channel Prior (MDCP), was proposed to improve underwater images based on the Dark Channel Prior algorithm and process the illumination component Y using the sigmoid function through the YIQ color space. Using non-reference quality measures Average Gradient (AG), CF, and Natural Image Quality Evaluator (NIQE), the proposed method was compared with other methods (UCM, RSU, RSAP, and MSRCR). The results showed the success of the proposed method in improving underwater images and they have good values.