论文部分内容阅读
对曲波变换的地震数据随机噪声衰减方法进行了探索,基于曲波(Curvelet)变换在图像处理方面的优越性,结合循环平移(Cycle spinning)技术提出了一种用于地震随机噪声衰减的新方法.在利用曲波变换闽值去噪算法基础上引入循环平移技术,可以消除曲波变换由于缺乏平移不变性所导致的信号伪吉布斯效应,并且较好地保留了有效信号.对地震正演模拟数据进行随机噪声衰减试验,对不同噪声含量数据进行去噪分析,并应用于实际地震数据,结果表明,新方法对去除地震随机噪音有较好的效果.
Based on the superiority of curvelet transform in image processing, this paper proposes a method of random noise attenuation of seismic data based on curvelet transform. Combining with Cycle spinning technique, a novel method for seismic random noise attenuation Method.Using Cyclone transform threshold denoising algorithm based on the introduction of cyclic translation technology can eliminate the pseudo-Gibbs effect caused by the lack of translational invariance of the curvelet transform, and better retain the effective signal.For seismic Random noise attenuation test was conducted on the forward modeling data. The noise data of different noise levels were denoised and applied to the actual seismic data. The results show that the new method is effective in removing random noise from earthquakes.