Environmental effects on seabed object bistatic scattering classification.

Area of Science:

• Geophysics
• Acoustics
• Underwater target detection

Background:

• Bistatic scattering from seabed targets is influenced by bottom type, affecting target classification accuracy.
• Inaccurate models neglecting bottom composition can lead to mischaracterization of target type, geometry, or material.

Purpose of the Study:

• To investigate the impact of bottom composition and self-burial on bistatic scattering from spherical and cylindrical targets.
• To compare experimental scattering data with simulation models incorporating varied bottom compositions and burial depths.
• To assess the robustness of target classification models to environmental variations.

Main Methods:

• Conducted autonomous underwater vehicle-based bistatic scattering experiments in a 6.5m deep environment with mud and sand bottoms.
• Developed and utilized scattering simulation models with diverse sediment parameters and target burial increments.
• Employed machine learning methods to train and test classification models using both simulated and experimental data.

Main Results:

• Correlations between experimental and simulated data were analyzed to validate model accuracy across different bottom compositions.
• Classification models trained on simulated data demonstrated robustness when applied to experimental data.
• Combined-environment classification models, considering mud depth and burial, proved effective for target classification.

Conclusions:

• Bottom type and target self-burial are critical factors in bistatic scattering analysis.
• Accurate modeling of sediment properties and burial is essential for reliable underwater target classification.
• Machine learning approaches incorporating environmental variability enhance the robustness of seabed target detection systems.