Underwater Image Enhancement Based on Multi-scale Exposure and Edge Protection

Underwater image plays an important role in exploitation of ocean resources. To tackle the problems of low contrast, blurred texture and serious color deviation in underwater images, considering the attenuation characteristics of light under water, we propose an underwater image enhancement algorithm based on multi-scale exposure and edge protection. Firstly, the original underwater image is processed by color pre-balancing and gamma transformation to generate multiple pseudo-exposure image sequences. The multi-scale exposure method, which combines global image information with local exposure characteristics, is used for preliminary contrast enhancement. Then, Retinex is used for color correction to obtain contrast-enhanced images. Next, the edge texture map and the maximum light energy loss map of the original underwater image are extracted separately. Finally, a fusion strategy based on edge protection and light energy loss is adopted to obtain the final enhanced underwater image. The experimental results show that the subjective and objective quality of the underwater enhanced images processed by the proposed method is better than that of the nine comparison methods, and is closer to natural scene images in terms of visual perception. The underwater image quality evaluation metrics UIQM and UCIQE reach 4.5946 and 0.6260 respectively.