Underwater image color correct in extremely poor visibility

In this paper, a new approach to realize underwater image color correction is proposed. The key problem of low visibility underwater vision is backscatter noise and color degradation. In order to solve this problem, we use color image histogram equalization to process the images taken by inhomogeneous illumination which intensity distribution was an inverse form of light attenuation pattern. The image obtained from turbid water become clear owing to inhomogeneous illumination and become colorful owing to color image histogram equalization which separate three color channels into three gray images. Experiment result decrease the backscattering effect and get a color-corrected image using this approach.