GPR signal de-noise method based on variational mode decomposition

TL;DR

This paper introduces a variational mode decomposition-based method for de-noising ground penetrating radar signals, demonstrating superior noise removal and high signal-to-noise ratios compared to traditional techniques.

Contribution

The paper applies VMD to GPR signal de-noising, providing a more efficient and theoretically grounded alternative to existing methods like wavelet transform and EEMD.

Findings

Effective noise removal in GPR signals

High SNR achieved under strong noise conditions

VMD outperforms WT and EEMD in experiments

Abstract

Compared with traditional empirical mode decomposition (EMD) methods, variational mode decomposition (VMD) has strong theoretical foundation and high operational efficiency. The VMD method is introduced to ground penetrating radar (GPR) signal processing. The characteristics of GPR signals validate the method of signal de-noising based on the VMD principle. The validity and accuracy of the method are further verified via Ricker wavelet and forward model GPR de-noising experiments. The method of VMD is evaluated in comparison with traditional wavelet transform (WT) and EEMD (ensemble EMD) methods. The method is subsequently used to analyze a GPR signal from a practical engineering case. The results show that the method can effectively remove the noise in the GPR data, and can obtain high signal-to-noise ratios (SNR) even under strong background noise.