Adaptive Real-Time Video Transmission through Underwater Acoustic Communication Link

Efficient video transmission enables a wide range of applications in underwater environments, such as seabed survey, subsea equipment maintenance, oil pipe/bridge inspection, and marine life sample collection. At present, it is a common belief that real-time underwater video transmission through underwater acoustic communication is challenging due to the influence of complex underwater environments and the limitation of underwater acoustic communication. In this paper, we propose an adaptive real-time underwater video transmission system using underwater communication. The system consists of three modules, i.e, video pre-processing module, video transmission module and video post-processing module. In the first two modules, the sender adaptively adjusts the compression bitrate and transmission rate according to the video quality and channel conditions. In the third module, the deep learning-based video reconstruction algorithm for underwater image information recovery is exploited. The efficacy of this system is verified by real underwater videos collected in several sea fields. The results prove the proposed system is able to transmit video successfully and efficiently in the underwater environment.