Curvelet-transform-based Random Noise Attenuation Using Energy Mean Scanning in f-k Domain

The curvelet denoising method has been successfully applied to seismic random noise attenuation. However, The conventional curvelet denoising method based on thresholding algorithm fails to preserve weak reflected signals in seismic events which have complex geometric structure. In this paper,we propose a novel curvelet denosing algorithm that applies the curvelet transform to 2D Fourier-transformed data in the fk domain instead of original data in the tx domain. In f-k domain in which the energy of the events with similar moveouts and diverse amplitude is located closely. Thus, we can take advantage of the energetic aggregation to preserve the weak reflected signals. Furthermore, an energy mean scanning algorithm based on the energetic statistical difference between effective signal and noise is proposed to perserve effective signal. The application on synthetic shot record and real imagration data demonstrates the feasibility and effectiveness of our proposed method.