Effective monitoring of marine activities is essential for optimizing the marine operations, ensuring the safety and protecting the marine environment. However, the existing monitoring approaches still have the problems of low intelligence, high cost and poor reliability. This article attempts to employ autonomous underwater vehicles (AUVs) to develop a ubiquitous monitoring system for marine activities. Particularly, it mainly includes AUVs, human operator, onshore cameras, and wireless communication networks. With the wireless communication networks, the human operator in the control center assigns monitoring tasks to AUVs and onshore cameras. After that, the cameras mounted on AUVs and shore collect the images of surface target (e.g., ship or buoy), such that the target attitudes can be estimated by using deep learning technology. For underwater target (e.g., shipwreck or drowning person), sonars are mounted on AUVs to carry out water depth detection. Finally, experimental results are provided to verify the effectiveness of our system. These results provide evidence that the AUV-assisted monitoring system can improve the reliability and reduce the cost over the manual monitoring system.
Yan et al. (Sat,) studied this question.