Underwater Video Enhancement Using Improved Retinex Algorithm

Underwater videos have different difficulties because of the influence of the surroundings on visibility and light. The light is not concentrated due to suspended particulates in the water, which gives a blurry appearance and reduced contrast. When the visual information is lowered, the quality significantly decreases, which is essential for many underwater uses, including surveillance, marine biological research, and underwater exploration. The main goal is to produce an enhanced underwater video that is useful in marine applications. In this paper, the Single Scale Retinex (SSR) algorithm is combined along with the conversion of RGB (Red, Green, Blue) to HSV (Hue Saturation Value) color space. This method enhances contrast and makes color manipulation easier by separating the hue from the value. The recommended technique starts by breaking up the underwater video into multiple frames and then converting each frame from RGB to HSV color space for further processing. This conversion enhances visibility and corrects color distortions more easily, isolating the illuminance component. The proposed method has shown much better results than existing algorithms.