A Compressive Multi-Frequency Linear Sampling Method for Underwater Acoustic Imaging.

Area of Science:

• Underwater acoustics
• Inverse scattering theory
• Signal processing

Background:

• The linear sampling method (LSM) is a qualitative inverse scattering technique used for imaging.
• LSM relies on solving unstable integral equations, which are further destabilized by under-sampled aperture data, especially in underwater scenarios.
• Limited aperture receiver configurations pose significant challenges for accurate underwater imaging.

Purpose of the Study:

• To propose an iterative inversion method for underwater imaging using limited aperture data.
• To enhance the linear sampling method (LSM) by incorporating a compressive sensing framework.
• To leverage multi-frequency data diversity for improved imaging stability and resolution.

Main Methods:

• Development of an iterative inversion method based on a compressive sensing framework.
• Application of multi-frequency data diversity by imposing a partial frequency variation prior.
• Formulation of an alternating direction method of multipliers (ADMM) to minimize the cost function.

Main Results:

• Demonstrated proof of concept using numerically generated data.
• Validated the method with experimental acoustic measurements in a shallow pool facility.
• Showcased the effectiveness of the multi-frequency approach for under-sampled aperture data.

Conclusions:

• The proposed iterative inversion method enhances LSM for underwater imaging with limited aperture data.
• Leveraging multi-frequency diversity within a compressive sensing framework improves imaging stability.
• The method shows promise for practical applications in underwater acoustic imaging.