基于U-Net的盐体识别方法

doi:10.19596/j.cnki.1001-246x.8044

计算物理 ›› 2020, Vol. 37 ›› Issue (3): 327-334.DOI: 10.19596/j.cnki.1001-246x.8044

基于U-Net的盐体识别方法

卢新瑞, 黄捍东, 李帅, 尹龙

中国石油大学(北京) 非常规油气科学技术研究院, 北京 102249

收稿日期:2019-01-21 修回日期:2019-05-17 出版日期:2020-05-25 发布日期:2020-05-25
作者简介:卢新瑞(1995-),硕士研究生,研究方向为复杂储层预测及深度学习在地震勘探中的应用,E-mail:rickyxinrui@163.com

Salt-body Classification Method Based on U-Net

LU Xinrui, HUANG Handong, LI Shuai, YIN Long

Unconventional Oil and Gas Institute of Science and Technology, China University of Petroleum(Beijing), Beijing 102249, China

Received:2019-01-21 Revised:2019-05-17 Online:2020-05-25 Published:2020-05-25

摘要/Abstract

摘要： 卷积神经网络在计算机视觉领域取得重大突破，利用其强大的图像处理能力，将地下沉积盐体的识别问题转化为图像语义分割问题，应用深度卷积神经网络实现盐体地震图像的像素级语义分割.本文在U-Net基础上，增加网络深度并同时引入批归一化和Dropout处理，使得神经网络模型具有更高的可信度和更强的泛化能力.通过实验发现，在卷积层之后引入批归一化处理，并在池化层和叠加层之后引入Dropout可以稳定提升模型对盐体图像的分割性能.

关键词: U-Net, 卷积神经网络, 盐体识别, 图像语义分割

Abstract: Convolutional neural network has made great breakthroughs in the field of computer vision. With its powerful image processing ability,we transform classification of underground sedimentary salt-body into image semantics segmentation problem. Deep convolution neural network is applied to implement pixel-level semantics segmentation of salt seismic images. We increase depth of the network and adds batch normalization and Dropout processing based on U-Net, which makes the neural network model with higher reliability and stronger generalization ability. With experiments, it was found that adding batch normalization layer after convolution layer, and adding Dropout after the pooling layer and concatenate layer improve steadily segmentation performance of the model for salt-body seismic image.

Key words: U-Net, convolution neural network, salt-body classification, image semantics segmentation

中图分类号:

P631.4

卢新瑞, 黄捍东, 李帅, 尹龙. 基于U-Net的盐体识别方法[J]. 计算物理, 2020, 37(3): 327-334.

LU Xinrui, HUANG Handong, LI Shuai, YIN Long. Salt-body Classification Method Based on U-Net[J]. CHINESE JOURNAL OF COMPUTATIONAL PHYSICS, 2020, 37(3): 327-334.

参考文献

[1] 唐祥华.世界含盐盆地中的油气资源[J]. 中国地质, 1990, (7):27-28.
[2] 王建谋. 盐丘检测[J]. 石油物探译丛, 1993, (2):25-29.
[3] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521:436-444.
[4] ZHOU Y H, LIU H Q, QI P, et al. Forecast of oil production in fractured-vuggy reservoir by using recurrent neural networks[J]. Chinese Journal of Computational Physics, 2018, 35(6):668-674.
[5] ZHANG Q F, YAO J, HUANG Z Q, et al. A multiscale deep learning model for fractured porous media[J]. Chinese Journal of Computational Physics, 2019, 36(6):665-672.
[6] GOU X Y, GUO L X, ZHANG L B. Support vector machine and neural network in inversion of rough surface parameters[J]. Chinese Journal of Computational Physics, 2014, 31(1):75-84.
[7] DI H, WANG Z, ALREGIB G. Deep convolutional neural networks for seismic salt-body delineation[C]. AAPG Annual Convention and Exhibition, 2018.
[8] SHI Yunzhi, WU Xinming, FOMEL Sergey. Automatic salt-body classification using deep-convolutional neural network[C]. SEG Technical Program Expanded Abstracts, 2018.
[9] LECUN Y, BOSER B, DENKER J S, et al. Backpropagation applied to handwritten zip code recognition[J]. Neural Computation, 1989, 1(4):541-551.
[10] LECUN Y, BENGIO Y. Convolutional networks for images, speech, and time series[J]. The Handbook of Brain Theory and Neural Networks, 1995, 3361(10):1995.
[11] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C]. Advances in Neural Information Processing Systems, 2012.
[12] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015.
[13] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016:770-778.
[14] RONNEBERGER O, FISCHER P, BROX T. U-net:Convolutional networks for biomedical image segmentation[C]. International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer, Cham, 2015.
[15] GOODFELLOW I, BENGIO Y, COURVILLE A, et al. Deep learning[M]. Cambridge:MIT Press, 2016.
[16] IOFFE S, SZEGEDY C. Batch normalization:Accelerating deep network training by reducing internal covariate shift[J]. arXiv preprint arXiv:1502.03167, 2015.
[17] NOH H, HONG S, HAN B. Learning deconvolution network for semantic segmentation[C]. Proceedings of the IEEE International Conference on Computer Vision, 2015.
[18] SHEKHOVTSOV A, FLACH B. Stochastic normalizations as Bayesian learning[J]. arXiv Preprint arXiv:1811.00639, 2018.
[19] SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al. Dropout:A simple way to prevent neural networks from overfitting[J]. The Journal of Machine Learning Research, 2014, 15(1):1929-1958.
[20] LI X, CHEN S, HU X, et al. Understanding the disharmony between dropout and batch normalization by variance shift[J]. arXiv Preprint arXiv:1801.05134, 2018.
[21] HINTON G E, SRIVASTAVA N, KRIZHEVSKY A, et al. Improving neural networks by preventing co-adaptation of feature detectors[J]. arXiv Preprint arXiv:1207.0580,2012.
[22] ZAGORUYKO S, KOMODAKIS N. Wide residual networks[J]. arXiv Preprint arXiv:1605.07146, 2016.
[23] IGLOVIKOV V, SHVETS A. TernausNet:U-Net with VGG11 encoder pre-trained on imageNet for image segmentation[J]. arXiv Preprint arXiv:1801.05746, 2018.
[24] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet classification with deep convolutional neural networks[C]//Advances in Neural Information Processing Systems, 2012:1097-1105.
[25] KINGMA D P, BA J. Adam:A method for stochastic optimization[J]. arXiv Preprint arXiv:1412.6980, 2014.

基于U-Net的盐体识别方法

Salt-body Classification Method Based on U-Net

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 4

编辑推荐

Metrics

作者中心

审稿中心

期刊浏览

期刊介绍

[1]	周枫林, 王炜佳, 廖海洋, 李光. 标量波传播问题的双互易精细积分法[J]. 计算物理, 2020, 37(1): 26-36.
[2]	宋林平. 阻尼加权最小二乘地震走时层析成象[J]. 计算物理, 1995, 12(4): 499-504.
[3]	王承曙, 李凤林. 褶积反演中的脉冲压缩准则[J]. 计算物理, 1994, 11(4): 498-514.
[4]	王承曙. 多道预测反褶积[J]. 计算物理, 1994, 11(3): 290-296.