Noise suppression based on a fast discrete curvelet transform

Abstract

A curvelet transform (CT) is a new kind of multi-scale analysis algorithm which is more suitable for image processing; as compared with a wavelet transform (WT), it can better analyse the line and curve edge characteristics, and it has better approximation precision and sparsity description and also has good directivity. This paper introduces ground roll suppression of a seismic signal based on a fast discrete curvelet transform (FDCT) and the direct wave removal in ground penetrating radar (GPR) based on a CT. A ground roll is an interferential wave existing widely in seismic data, and it is an important step in seismic data processing to suppress ground roll. The method can separate the different waves at the same time because of CT characteristics of frequency and direction. In the ground penetrating radar signal, the direct wave is a strong noise which has line singularity. In the time-space domain, the useful information in the GPR signal is badly polluted by a strong direct wave so that the underground objects cannot be recognized correctly. The FDCT is employed to remove the background noise and separate the direct wave. The experimental results show that a CT has better effects in ground-roll suppression and direct wave removal.