Problem-based UAV Design for Port Surveillance and Monitoring and Machine Learning for Detection and Identification of Vessels

This paper presents a system design methodology for UAV-based surveillance to monitor vessel traffic and to detect encroachments in port areas. Different solutions in terms of types and configurations of UAVs are proposed for different applications and problems. Though there are similarities in solution approaches, the environment and constraints posed by each problem may not allow direct use of one solution to the other. Therefore, a holistic solution may be needed for the problem such as port surveillance, vessel detection and encroachment. In this work, a systematic approach is proposed for the design of a UAV, and simulation of machine learning for vessel detection and identification is demonstrated.