Author(s)

Yiheng Wu, XinLi, Zhangjie Yin, Jian Li and Chun Liu

Corresponding Author

Chun Liu

ABSTRACT

Based on deep learning, the design of smart navigation mark light system proposes to extend traditional navigation mark light as an unattended platform for waterway. While actively acquiring the working status of navigation mark light and timely feeding back the damage information to the back-end server, the system can also monitor the passing vessels on waterway at all times. In view of the main cause of damage to navigation mark light for ship impact, this system can retain evidence and warn the management so the economic loss caused by hit-and-run vessels. After real-scene experiments on the Yangtze River, the navigation mark light equipped with this system can be used as an unattended platform and implement all functions.

KEYWORDS

Deep learning, Smart navigation mark light, MobileNet-SSD, Waterway monitoring