A Novel Method Based on Proximate Wavelet Coefficient Recovery for Magnetic Resonance Sounding Signal Denoising in Complex Interference Environments.

Magnetic resonance sounding (MRS) is the only technology capable of noninvasive direct detection of subsurface water content. However, MRS often suffers from weak signals (10 ^-9 V) and may prohibit application in complex interference environments. To effectively recover the MRS signals regardless of whether they are corrupted by single spike or cascades of spikes, a method based on proximate wavelet coefficient recovery technology is proposed. Compared with the conventional methods, the proposed method improves the measurement efficiency by eliminating the demand for multiple recordings. The accuracy and stability of the proposed method are investigated using simulated signals from complex noise environments and geological conditions. Simulation results show significant improvements in signal-to-noise ratio (SNR) and retrieval of signal parameters, which demonstrate the validity of the proposed method. Moreover, the proposed method is applied on the synthetic signals embedded in noise-only data recorded in the urban environment. The results show that the proposed method can maintain a balance between preserving the signal component and suppressing complex spiky noise. The proposed method is implemented for the field data and the results demonstrate good practicality. The research results provide technical support for rapid detection in complex interference environments.