Why using CNN for seismic interpretation? An investigation
Why using CNN for seismic interpretation? An investigation
复制标题
为什么使用 CNN 进行地震解释？
DOI：
10.1190/segam2018-2997155.1
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
G. AlRegib
中科院分区：
文献类型：
--
作者：
H. Di;Zhen Wang;G. AlRegib
三维地震解释在地下储层的稳健油气勘探和生产中起着关键作用。然而，随着3D地震勘探规模的急剧增长，手动解释地震体变得更具挑战性。近年来，人工智能和机器学习技术已经成功地应用于各个学科，这极大地促进了其在地震领域的应用，以模仿有经验的解释者的智能并在各种任务中辅助地震解释，包括相分析和构造检测（例如，断层和盐丘）。在这项研究中，我们首先应用两个最流行的神经网络框架，多层感知器（MLP）网络和卷积神经网络（CNN），地震盐体划定的问题，并比较它们的性能。然后，我们研究了有助于CNN框架在理解地震信号和识别重要地震结构方面表现更好的两个因素。具体来说，一方面，CNN能够从原始地震图像自动生成一套特征，这减少了对解释器计算和调整地震属性的依赖。另一方面，更重要的是，CNN分类是基于块的，其中考虑了局部地震反射模式来定义和学习目标结构的特征。以这种方式，可以有效地识别和排除不同图案的随机/相干地震噪声和处理伪影。
Summary Three-dimensional seismic interpretation plays a key role in robust hydrocarbon exploration and production of subsurface reservoirs. With the dramatic growing size of 3D seismic surveys, however, manually interpreting a seismic volume turns to be even more challenging. In recent years, artificial intelligence and machine learning techniques have been successfully applied in various disciplines, which greatly promotes its applications in the seismic domain for mimicking an experienced interpreter’s intelligence and assisting seismic interpretation in various tasks, including facies analysis and structure detection (e.g., faults and salt domes). In this study, we first apply two most popular neural network frameworks, the multi-layer perceptron (MLP) network and the convolutional neural network (CNN), to the problem of seismic salt-body delineation and compare their performance. Then, we investigate two factors that contribute to the better performance of the CNN framework in understanding seismic signals and identifying the important seismic structures. Specifically, on one hand, the CNN is capable of automatically generating a suite of features from the original seismic images, which reduces the dependency on interpreters for computing and tuning seismic attributes. On the other hand and more importantly, the CNN classification is patch based, in which local seismic reflection patterns are taken into account for defining and learning the features of the target structures. In this way, the random/coherent seismic noise and processing artifacts of distinct patterns can be effectively identified and excluded.