A Multi-Feature and Multi-Level Matching Algorithm Using Aerial Image and AIS for Vessel Identification.

Area of Science:

• Maritime surveillance
• Robotics and Automation
• Data Fusion

Background:

• Effective monitoring and management of vessels in maritime channels are crucial for safety and security.
• Existing methods for vessel identification and tracking often face limitations in accuracy and autonomy.

Purpose of the Study:

• To develop an integrated maritime unmanned aerial vehicle (Mar-UAV) system for enhanced vessel monitoring.
• To propose a robust multi-feature and multi-level matching algorithm for accurate vessel detection and identification.

Main Methods:

• Development of a Mar-UAV system equipped with a high-resolution camera and Automatic Identification System (AIS).
• Implementation of a novel matching algorithm utilizing spatiotemporal characteristics from aerial imagery and AIS data.
• Division of the matching algorithm into point matching and trajectory matching for improved accuracy.

Main Results:

• The Mar-UAV system successfully integrated high-resolution imaging with AIS data for vessel identification.
• The multi-feature, multi-level matching algorithm demonstrated high accuracy in detecting and identifying vessels.
• Field experiments in the Yangzi River confirmed the system's effectiveness in autonomous maritime supervision.

Conclusions:

• The developed Mar-UAV system and its matching algorithm significantly improve autonomous vessel identification and tracking.
• This technology offers a substantial advancement for maritime supervision and management.
• The system enhances the autonomy of UAVs in performing complex maritime tasks.