Embedded Marine Target Detection and Positioning Shipboard Terminal Based on Beidou Navigation Satellite System

Improving the maritime information supervision system and establishing a strong maritime nation is one of the important strategies for national defense. Civilian fishing vessels, as the main body of marine economic activities, are important nodes for sensing marine information. In this paper, lightweight neural network is deployed to embedded devices for maritime target recognition. Target ranging is realized by building a maritime monocular visual ranging model. And the high precision real-time ship information of heading and speed is obtained by sensors. The positioning of the observation vessel is obtained based on BeiDou Navigation Satellite System (BDS). And the obtained maritime information is transmitted back to the shore terminal through Regional Short Message Communication (RSMC). The detection of maritime targets and the detection of the surrounding environment are realized through embedded devices using maritime vessel nodes. Monitoring of suspicious targets at sea and data return can be completed. During the simulation experiments, the relative error rate of ranging was below 18.2%.