Multistatic acoustic characterization of seabed targets

Area of Science:

• Underwater acoustics
• Robotics
• Oceanography

Background:

• Autonomous underwater vehicles (AUVs) are crucial for seabed object detection.
• Passive acoustic sensing by AUVs can discriminate targets based on scattered sound fields.
• Characterizing seabed objects is vital for safety and environmental monitoring.

Purpose of the Study:

• To investigate the use of acoustic scattering data from mobile sources and receivers for target characterization.
• To develop and evaluate a discrimination approach for classifying seabed objects using multistatic scattering.
• To assess the feasibility of using low-cost AUVs for hazardous object detection.

Main Methods:

• Utilized the OASES-SCATT simulator to model acoustic scattering from spherical and cylindrical targets.
• Explored the impact of target geometry on multistatic scattering fields.
• Developed a classification approach using frequency components of scattering data and relative angles between vehicles and target.

Main Results:

• Achieved classification accuracies greater than 98% for target shape and composition.
• Demonstrated potential for target volume regression with a 90% chance of errors less than 15%.
• Validated the effectiveness of the multistatic scattering approach for target discrimination.

Conclusions:

• Acoustic scattering analysis using mobile AUVs offers a viable method for seabed object classification.
• The developed approach enables accurate identification of target shape and composition.
• This technique supports the use of low-cost autonomous vehicles for underwater hazard detection.