Modified Restoration Technique for Improved Visual Perception of Shallow Underwater Imagery

Images captured underwater often suffer from quality degradation such as low contrast, non-uniform illumi-nation, etc. due to the attenuation and backscattering of light by suspended underwater particles. To over-come this, restoration-cum-enhancement techniques are necessary. Here we present the modified under-water light attenuation prior (MULAP) model using supervised linear regression model to restore the degraded image. The image formation model (IFM)-based restoration depends on dual factors: back-ground light and transmission map.Initially, datasets are collected on the close-range point-of-interest. Then, experimental analyses are carried out for those images using the IFM-based methods. For the above techniques, both subjective analysis and objective analysis are done by considering dual metrics such as universal quality index and visual information fidelity factor. Finally, the proposed MULAP shows over-whelming qualitative and quantitative results among other state-of-the-art techniques.