GPR Clutter Removal Based on Weighted Nuclear Norm Minimization for Nonparallel Cases

Area of Science:

• Geophysics
• Electromagnetic Methods
• Signal Processing

Background:

• Ground-penetrating radar (GPR) is crucial for underground target detection.
• Strong clutter often obscures target signals, degrading detection performance.
• Existing methods struggle with nonparallel antenna-ground surface conditions.

Purpose of the Study:

• To develop a novel GPR clutter-removal method for nonparallel antenna-ground scenarios.
• To improve the accuracy and efficiency of underground target detection using GPR.
• To address the limitations of current clutter-removal techniques.

Main Methods:

• Proposed a novel GPR clutter-removal method based on weighted nuclear norm minimization (WNNM).
• Decomposed GPR B-scan images into low-rank clutter and sparse target matrices.
• Utilized a non-convex weighted nuclear norm with differential singular value weighting.

Main Results:

• The WNNM method demonstrated superior performance in numerical simulations and real GPR experiments.
• Comparative analysis showed higher peak signal-to-noise ratio (PSNR) and improvement factor (IF) than state-of-the-art methods.
• The proposed method is approximately five times faster than Robust Principal Component Analysis (RPCA).

Conclusions:

• The WNNM method effectively removes clutter in nonparallel GPR scenarios.
• This technique significantly enhances underground target detection performance.
• The method's speed and effectiveness make it suitable for practical GPR applications.