Abstract
Underwater images exhibit poor visibility, such as color distortion, low contrast and blur, which results from exponential attenuation that the propagated light suffers in water. The degraded underwater images pose challenges to underwater target detection and further analysis. To address these problems, this paper presents a hybrid underwater image processing approach which consists of an underwater image restoration algorithm and a color balance algorithm. In our work, we design a novel scoring formula to accurately locate background region. Meanwhile, we innovatively decompose the attenuation curves on RGB channels to calculate transmission maps. The proposed algorithm can effectively estimate the transmission maps and the background light from a single underwater image, which can be applied to restore the underwater image. Moreover, a color balance algorithm is introduced to output an enhanced image with natural appearance. Ultimately, comprehensive tests are implemented to evaluate performance of our method. Experiments on images captured in different underwater scenes manifest that the proposed method outperforms several prominent underwater image enhancement and restoration approaches.
Accepted Version (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call