Cooperative unmanned surface vehicles and unmanned aerial vehicles platform as a tool for coastal monitoring activities

To ensure the safety of near-shore activities, efficient coastal monitoring equipment must be installed. Coastal monitoring has become a priority in integrated marine management because casualty accidents occur frequently in coastal waters. In this paper, the residual module of a CenterNet network is modified in response to the degradation problem. And by replacing the original convolutional layer with several depth-wise separable convolutions, the accuracy of our model's recognition performance for ships and personnel improved. Then an improved CenterNet algorithm, which has a 4% increase in average precision over the original one, is adopted to identify vessels and participants at sea in the images returned from the cooperative unmanned surface vehicles(USV) and unmanned aerial vehicles(UAV) platform. A cooperative USV-UAV platform for coastal surveillance is built by merging with a CenterNet-improved identification network. Correspondingly, several related management practices and potential application areas are recommended on the basis of the platform. It provides a framework for inshore safety rescue operations as a distinct cooperative USV-UAV platform.