Direct‐wave denoising of low‐frequency ground‐penetrating radar in open pits based on empirical curvelet transform

Abstract

Low‐frequency ground‐penetrating radar is known to be effective in the detection of geological disasters in open pits owing to its good detection depth, high resolution and portability. However, because of the lack of a shield layer and the poor anti‐interference capability, low‐frequency ground‐penetrating radar has certain limitations. A strong linear interference due to direct waves can result in a complex mining environment, hindering the detection of geological disasters in open pits. Conventional methods, such as wavelet transform and curvelet transform, are ineffective, as they cannot adaptively remove noise in accordance with signal characteristics. In this paper, considering the high apparent velocity and energy of direct‐wave noise in open‐pit ground‐penetrating radar data, an empirical curvelet transform method is proposed to suppress the interference signals. The empirical curvelet transform not only is multi‐scale and multi‐directional, but can also adaptively perform filtering based on the characteristics of the ground‐penetrating radar interference signals. This method is applied to the simulation and processing of measured ground‐penetrating radar data and is compared with the conventional curvelet transform. The results confirm the effectiveness of this method.

Deep Convolutional Denoising Autoencoders with Network Structure Optimization for the High-Fidelity Attenuation of Random GPR Noise

“…However, because of strong linear interference, the conventional curvelet transform is ineffective for noise removal in this case, because it cannot adaptively remove noise according to the signal characteristics; hence, a method, called the empirical curve wave transform, which can suppress interference signals, was proposed and compared with the conventional curvelet transform. The results confirmed the effectiveness of the method [9]. In addition, to remove noise from GPR echo signals, a denoising method that was based on ensemble EMD (EEMD) and the wavelet transform was presented in [10]; as compared with other common methods, the EEMD-wavelet method improves the SNR.…”

A ground penetrating radar denoising method based on CEEMD and wavelet decomposition

“…Due to the use of a wide range of frequencies, the influence of field exploration conditions and interfere of surrounding electromagnetic waves, the recorded signal contains a lot of noise. Especially, when using lowfrequency radar, although its detection depth is large, but because of the lack of a shield layer and the poor anti-interference capability, the amount of noise is further increased (T. He & Shang, 2020). In addition, the manual data interpretation results largely depend on the experience and knowledge level of the interpreter, so the advanced applications of GPR require engagement of skilled, trained and experienced professional (Wai-Lok Lai, Dérobert, & Annan, 2018).…”