摘要：The Radon transform is widely used for multiple elimination. Since the Radon transform is not an orthogonal transform, it cannot preserve the amplitude of primary reflections well. The prediction and adaptive subtraction method is another widely used approach for multiple attenuation, which demands that the primaries are orthogonal with the multiples. However, the orthogonality assumption is not true for non-stationary field seismic data. In this paper, the high-order sparse Radon transform (HOSRT) method is introduced to protect the amplitude variation with offset information during the multiple subtraction procedures. The HOSRT incorporates the high-resolution Radon transform with the orthogonal polynomial transform. Because the Radon transform contains the trajectory information of seismic events and the orthogonal polynomial transform contains the amplitude variation information of seismic events, their combination constructs an overcomplete transform and obtains the benefits of both the high-resolution property of the Radon transform and the amplitude preservation of the orthogonal polynomial transform. A fast nonlinear filter is adopted in the adaptive subtraction step in order to avoid the orthogonality assumption that is used in traditional adaptive subtraction methods. The application of the proposed approach to synthetic and field data examples shows that the proposed method can improve the separation performance by preserving more useful energy.

关键字：multiple attenuation Radon transform orthogonal polynomial transform high-order sparse Radon transform overcomplete dictionary

ISSN号：1742-2132

卷、期、页：卷: 13 期: 3 页: 207-219

发表日期：2016-06-01

期刊分区（SCI为中科院分区）：四区

收录情况：SCIE（科学引文索引网络版）,ESI（基本科学指标数据库）

发表期刊名称：JOURNAL OF GEOPHYSICS AND ENGINEERING

参与作者：Chen, Yangkang

通讯作者：杨静

第一作者：薛亚茹,马继涛

论文类型：期刊论文

论文概要：薛亚茹,杨静,马继涛,Chen, Yangkang，Amplitude-preserving nonlinear adaptive multiple attenuation using the high-order sparse Radon transform，JOURNAL OF GEOPHYSICS AND ENGINEERING，2016，卷: 13 期: 3 页: 207-219

论文题目：Amplitude-preserving nonlinear adaptive multiple attenuation using the high-order sparse Radon transform