Deep Sea Image Enhancement Method Based on the Active Illumination

To solve the low contrast and color distortion problem of deep sea image caused by active light scattering and absorption effects in the underwater environment, an underwater image enhancement method is proposed. Different from the previous methods, which estimate the background light with the brightest pixels, background light is estimated based on the non-correlation of the object and the background light, to alleviate the disturbance of the pixels in the white objects or the illuminated foreground region, while keeping its accuracy in removing scattering, and improve the underwater image contrast. Aiming at the color distortion caused by the color gain of artificial light source color and the optical attenuation, the grey pixels, which are close to the light source, are picked in the dehazed image. Then the light intensity can be derived with the detected pixels according to the sensitivity to the source. With the estimated light intensity, the light source color is achievable. At last, color distortion can be corrected by removing the source color while compensating for the optical attenuation. Experimental results demonstrate the proposed method can effectively remove haze, recover the relatively genuine color, and further obtain the enhanced image. The information entropy and the underwater image quality evaluation values of the proposed method are higher than that of the existing methods, which indicates that the proposed method can improve the underwater image quality significantly while preserving the efficient information.