Underwater Multi-Vehicle Trajectory Alignment and Mapping Using Acoustic and Optical Constraints

Area of Science:

• Robotics
• Underwater Systems
• Computer Vision

Background:

• Multi-robot formations enhance survey capabilities by merging individual robot capacities.
• Cooperative navigation of underwater vehicles often relies on control-level directives and range measurements, leading to a lack of global positioning awareness.
• Existing methods for multi-vehicle data processing are hindered by the absence of precise global positioning information.

Purpose of the Study:

• To present a global alignment method for correcting dead reckoning trajectories of multiple underwater vehicles using inter-vehicle acoustic messages.
• To extend an optimization framework for multi-vehicle geo-referenced optical 3D mapping using monocular cameras.
• To improve the computational efficiency of optical mapping by integrating optical constraints without traditional bundle adjustment.

Main Methods:

• A novel global alignment method is proposed to refine vehicle trajectories based on acoustic communication data.
• An extended optimization framework fuses optical reconstruction data with navigation information for 3D mapping.
• Optical constraints are incorporated in a computationally efficient manner, differing from standard bundle adjustment techniques.

Main Results:

• The proposed global alignment method successfully corrects dead reckoning trajectories of multiple vehicles.
• The enhanced optimization framework enables efficient multi-vehicle geo-referenced optical 3D mapping.
• The method demonstrates effective fusion of optical and navigation data for improved mapping accuracy.

Conclusions:

• The developed methods enhance the data processing capabilities of multi-robot formations, particularly for underwater surveys.
• The approach provides a generic and computationally efficient process for fusing optical and navigation data in multi-vehicle systems.
• The performance is validated using real-world datasets from the Morph EU-FP7 project, confirming its practical applicability.