Single Underwater Image Enhancement via Contrast Stretching and Fusion

In this paper, a novel single underwater image enhancement method is proposed to address color distortion, poor contrast and detail blur in underwater degraded images. First, the two attenuation channels of underwater degraded images are color compensated in the RGB color space. Then, the improved White Patch hypothesis is used to implement color correction. Moreover, contrast stretching and detail preservation of color-corrected images are carried out in the Lab color space. Finally, the two images and the weights defined are fused to obtain the final image at multiple scales. The experiment shows that the proposed method can effectively improve color distortion, enhance contrast and reduce blur in multiple underwater scenes, which is superior to existing methods.