Fault Detection and Characterization Using a 3D Multidirectional Sobel Filter

doi:10.2118/168061-ms

Fault Detection and Characterization Using a 3D Multidirectional Sobel Filter

2013-05-19

0

PRE

AI

DOI：10.2118/168061-ms

原文链接

原文求助

摘要

En 中文

Abstract The Sobel filter is a discrete differentiation operator widely used in seismic image processing algorithms for automatic fault detection and extraction. The filter approximates the local gradient by combining derivatives of the amplitude between neighbouring traces along the x, y, and z directions. However, the approximation is not accurate in noisy seismic data. Also, in areas of steeply dipping geological faults the approximation can lead to misidentification of the faults. To overcome these problems while also improving automatic fault detection, we propose herein a novel 3D Multidirectional Sobel Filter in which derivatives along nine non-orthogonal directions are computed, and then the direction with the largest derivative is chosen as the local gradient. This new 3D Multidirectional Sobel Filter is computationally intensive; therefore we leveraged a multi-core parallel computing design to reduce total execution time in our tests. We produced a speedup of 4x on an 8-core CPU using a 400 GB poststack seismic volume. Tests performed on synthetic and Saudi Arabian field data demonstrate that our 3D Multidirectional Sobel Filter delineates fault and fracture distributions more effectively than the conventional Sobel filter. Moreover, our multidirectional filter is shown to be less susceptible to acquisition and processing artefacts than the conventional Sobel filter.

Fault Detection and Characterization Using a 3D Multidirectional Sobel Filter

摘要

AI总结

期刊

机构

引用论文