An Improved Method for Underwater Image Super-Resolution and Enhancement

Underwater object detection and recognition are still challenging because of the degradation caused by the complex underwater imaging environment and lighting conditions. Aiming at the problems of the simultaneous enhancement and super-resolution (SESR), a method with single image super-resolution and enhancement of underwater imagery is proposed. Firstly, an underwater image super-resolution algorithm is proposed to remove blurring, and then an end-to-end underwater image enhancement network is used to compensate for color casts and produce natural color enhancement images. In addition, we verify the effectiveness of the proposed model through qualitative analysis and quantitative experiments and compare the performance with several state-of-the-art models, which show that the proposed method performs better.