Close-Range Tracking of Underwater Vehicles Using Light Beacons

Area of Science:

• Robotics
• Computer Vision
• Marine Technology

Background:

• Acoustic localization is limited for short-range, high-accuracy needs in autonomous underwater vehicle (AUV) formations.
• Safe and efficient close-formation navigation requires precise, high-update-rate positioning.

Purpose of the Study:

• To develop and evaluate a novel computer vision-based tracking system for AUVs in close formations.
• To provide accurate and high-speed pose estimation for target vehicles at short ranges.

Main Methods:

• Utilized computer vision techniques with active light markers for underwater vehicle tracking.
• Employed an omnidirectional camera for a full lower hemisphere field of view.
• Tested the system in real sea conditions during AUV mapping missions.

Main Results:

• The system accurately estimates the pose of target vehicles at short ranges.
• Achieved high execution speed for real-time tracking.
• Demonstrated robust operation over extended periods in real-world scenarios.

Conclusions:

• The proposed system offers a viable solution for precise, short-range AUV tracking, overcoming limitations of acoustic systems.
• The use of an omnidirectional camera enables multi-vehicle tracking and enhances situational awareness.
• The system's robustness in real sea conditions validates its effectiveness for underwater navigation.