Backscattering Analysis at ATR on Rough Surfaces by Ground-Based Polarimetric Radar Using Coherent Decomposition

Area of Science:

• Radar remote sensing
• Electromagnetic wave propagation
• Automatic Target Recognition (ATR)

Background:

• Backscattering analysis is crucial for ground-based radar systems, especially in complex environments like rough terrain.
• Wave polarization properties significantly influence radar signal characteristics and target detection capabilities.
• Existing decomposition theorems provide frameworks for analyzing polarized reflected waves.

Purpose of the Study:

• To compare the effectiveness of linear and circular polarized waves in radar backscattering analysis for automatic target recognition (ATR).
• To evaluate coherent decomposition methods, specifically Pauli, Krogager, and Cameron, under rough terrain conditions.
• To demonstrate an approach for extracting and utilizing polarization signal backscattering data for vehicle classification.

Main Methods:

• Utilized coherent decomposition theorems (Pauli, Krogager, Cameron) to analyze backscattered waves from ground-based radar.
• Extracted polarization signal backscattering data for two distinct vehicle types on rough terrain.
• Employed a corner reflector for calibrating measurement results, accounting for ground surface properties, and applied supervised learning algorithms for ATR.

Main Results:

• Pauli decomposition showed advantages for analyzing radar backscattering on rough terrain.
• Object classification accuracy reached 68.1% with linearly polarized waves and 54.2% with elliptically polarized waves.
• Reflected depolarized waves showed significant structural changes due to object rotation and rough surface influences, impacting recognition accuracy.

Conclusions:

• Linear polarization is recommended as a superior mechanism for object recognition in ATR systems operating on rough terrain.
• The influence of rough surfaces and object profile rotation degrades recognition accuracy when using depolarized waves.
• Calibrated backscattering data, combined with supervised learning, offers a viable path for ATR, with polarization choice being critical.