A Method of Image Enhancement in Shallow Water with Underwater System

Abstract

Absorption and scattering are two key issues of underwater optical imaging, which lead to fogging and color deviation. To solve this problem, a shallow sea image enhancement method with underwater imaging system is proposed. Firstly, taking the 3C-HF (Color Channel Compensation-Homomorphic Filtering) algorithm aims to remove the image uneven and dimness, which equilibrate with the original image from underwater equipment. Secondly, the preliminary enhancement images perform gamma correction and contrast limited adaptive histogram equalization (CLAHE) to become the fusion image. Meanwhile, the weights of these fusion images are defined to reduce the influence of edge, color and contrast on image fusion. To avoid the artifacts in the low frequency component of the reconstructed image caused by the sharp weighted image transition, a multi-scale fusion method is adopted. This method can not only improve the contrast of underwater images and reduce the problem of scattering and blurring, but also overcome the problem of color distortion. As experiments shown, compared with the state-of-the-arts, it not only can effectively realize the sharpness and color enhancement of various underwater images in a wide range, but also improve the completion of high-level tasks of underwater images.