| 基于Shearlet变换的尺度方向自适应阈值地震数据随机噪声压制方法 |
| |
| 引用本文： | 陈毅军,程浩,巩恩普,薛林.基于Shearlet变换的尺度方向自适应阈值地震数据随机噪声压制方法[J].吉林大学学报(地球科学版),2021,51(4):1231-1242. |
| |
| 作者姓名： | 陈毅军 程浩 巩恩普 薛林 |
| |
| 作者单位： | 深部金属矿山安全开采教育部重点实验室(东北大学), 沈阳 110819 |
| |
| 基金项目： | 国家自然科学基金项目（41804103）；国家重点研发计划项目（2017YFC1503101） |
| |
| 摘 要： | 地震勘探的有效信号常受到随机噪声的干扰而难以识别，需要进行随机噪声和有效信号的分离。传统Shearlet全局阈值不随方向与尺度变化，在去噪的同时也会损失许多有效信号。Shearlet变换作为一种新的多尺度多方向时频分析方法，具有最优的稀疏表示能力、局部化特征和方向敏感性。本文将含噪地震信号通过Shearlet分解后计算各尺度与方向上Shearlet域系数的L2范数，并对其进行数据重排后发现，随着方向改变L2范数不断减小，进而提出一种基于L2范数的尺度方向自适应阈值计算方法。将其与小波变换、曲波变换、Shearlet全局阈值去噪方法对比，模拟数据与实际地震记录去噪结果表明，本文方法在去除随机噪声的同时，深部弱信号也得到了很好的恢复，地震数据的信噪比比其他3种方法有所提高，在0.929 9 dB条件下提升至11.565 1 dB。
|
| 关 键 词： | Shearlet 地震数据 阈值改进 信噪比 |
| 收稿时间： | 2020-05-04 |
Random Noise Suppression of Seismic Data with Scale-Oriented Adaptive Threshold Based on Shearlet Transform |
| |
| Chen Yijun,Cheng Hao,Gong Enpu,Xue Lin.Random Noise Suppression of Seismic Data with Scale-Oriented Adaptive Threshold Based on Shearlet Transform[J].Journal of Jilin Unviersity:Earth Science Edition,2021,51(4):1231-1242. |
| |
| Authors: | Chen Yijun Cheng Hao Gong Enpu Xue Lin |
| |
| Institution: | Key Laboratory of Safe Mining of Deep Metal Mines(Northeastern University), Ministry of Education, Shenyang 110819, China |
| |
| Abstract: | The effective signal of seismic prospecting is often interfered by random noise, which is difficult to identify. It is necessary to separate the random noise from the effective signal. The traditional Shearlet global threshold does not change with the direction and scale, which leads to the loss of many effective signals during denoising. As a new multi-scale and multi-directional time-frequency analysis method, Shearlet transform has the best sparse representation ability, local feature, and direction sensitivity. In this paper, the noisy seismic signal is decomposed by Shearlet to calculate the L2 norm of Shearlet domain coefficients in various scales and directions. After rearranging the data, it is found that the L2 norm decreases as the direction changes. Based on the L2 norm, a calculation method of adaptive threshold value of the scale direction of the number is proposed. Compared with wavelet transform, curvelet transform, and Shearlet global threshold denoising methods, the denoising results of simulated data and actual seismic records show that the method in this paper can remove the random noise and can also recover the weak deep signals well, and the signal noise ratio of the data is improved from 0.929 9 dB to 11.565 1 dB. |
| |
| Keywords: | Shearle seismic data threshold improvement signal noise ratio |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《吉林大学学报(地球科学版)》浏览原始摘要信息 |
| 点击此处可从《吉林大学学报(地球科学版)》下载免费的PDF全文 |
