Abstract:The traditional first arrival picking method cannot take into account picking effect and efficiency,the algorithm stability is poor,and the industrial application has not been very mature.The first arrival picking method based on deep learning is time-consuming and labor-intensive,the process of data preprocessing is cumbersome,and the network structure is too complex,resulting in low training and test efficiency.Combining the advantages of U-Net with those of SegNet,a new hybrid network U-SegNet is constructed,and based on which first arrivals can be picked automatically.Based on the SegNet structure,U-SegNet provides multi-scale information of the encoder network by fusing jump connections information before the deconvolution layer of the decoder network to obtain better performance,and its upsampling operation changes the deconvolution in U-Net to unpooling.Because the pooling index is passed to the upsampling layer,the network model converges faster.Therefore,the U-SegNet network structure is more conducive to segmenting the background noise area and the area where background noise and valid signal overlap,thereby improving the accuracy of first arrival picking.The first arrival automatic picking process based on U-SegNet includes making a training data set,designing a network model,training the network model,testing the network model and applying it to real seismic data.Tests and applications of the U-SegNet model show that the picking efficiency of the proposed method is about 2.2 times that of a commercial software.It is easy to be industrialized and has a good future in large-scale application.
陈金焕,曹永生,孙成龙,等.基于二分法的地震波初至自动拾取算法[J].地球物理学进展,2015,30(2):688-694.CHEN Jinhuan,CAO Yongsheng,SUN Chenglong,et al.The algorithm for automatic first-breaks picking on seismic traces based on Dichotomy[J].Progress in Geophysics,2015,30(2):688-694.
[2]
宋建国,李赋真,徐维秀,等.改进的神经网络级联相关算法及其在初至拾取中的应用[J].石油地球物理勘探,2018,53(1):8-16.SONG Jianguo,LI Fuzhen,XU Weixiu,et al.An improved neural-network cascade-correlation algorithm and its application in seismic first break picking[J].Oil Geophysical Prospecting,2018,53(1):8-16.
[3]
林凡生,邹志辉.基于波形相似性的远震初至拾取方法进展与对比研究[J].地球物理学进展,2017,32(4):1597-1606.LIN Fansheng,ZOU Zhihui.Review and comparative study of waveform-similarity based teleseismic first-arrival picking methods[J].Progress in Geophysics,2017,32(4):1597-1606.
[4]
江国明,张贵宾,徐峣.远震相对走时数据快速计算方法及应用[J].地球物理学报,2012,55(12):4097-4105.JIANG Guoming,ZHANG Guibin,XU Yao.A fast method for calculating relative residuals of teleseismic data and its application[J].Chinese Journal of Geophysics,2012,55(12):4097-4105.
[5]
许鑫,张凯,王非翊,等.基于地震DNA算法的一种新的初至拾取方法[J].地球物理学进展,2020,35(5):1894-1901.XU Xin,ZHANG Kai,WANG Feiyi,et al.A new method of first break picking based on seismic DNA algorithm[J].Progress in Geophysics,2020,35(5):1894-1901.
[6]
Coppens F.First arrival picking on common-offset trace collections for automatic estimation of static corrections[J].Geophysical Prospecting,1985,33(8):1212-1231.
[7]
孙淑琴,王彦春,何巍巍,等.基于地震时窗属性特征拾取初至方法对比研究[J].勘探地球物理进展,2009,32(6):441-444.SUN Shuqin,WANG Yanchun,HE Weiwei,et al.Comparative study of window attribute based first arrival picking methods[J].Progress in Exploration Geo-physics,2009,32(6):441-444.
[8]
桑运云.初至波走时层析速度建模方法研究[D].山东青岛:中国石油大学(华东),2013.SANG Yunyun.Research on Modeling Method of Tomographic Velocity in First Arrival Wave[D].China University of Petroleum (East China),Qingdao,Shandong,2013.
[9]
祝佰航.三维VTI介质初至波走时层析速度和各向异性参数建模方法研究[C].CPS/SEG北京2018国际地球物理会议暨展览电子论文集,2018,481-484.ZHU Baihang.Research on modeling method of tomographic velocity and anisotropy parameters for first-time wave travel in 3D VTI media[C].CPS/SEG Beijing 2018 International Geophysical Conference and Exhibition Electronic Proceedings,2018,481-484.
[10]
潘树林,高磊,陈辉,等.可控震源地震记录初至拾取方法研究[J].石油物探,2010,49(2):209-212.PAN Shulin,GAO Lei,CHEN Hui,et al.Study on methods for picking up first break of seismic record acquired by vibroseis[J].Geophysical Prospecting for Petroleum,2010,49(2):209-212.
[11]
潘树林,高磊,邹强,等.一种实现初至波自动拾取的方法[J].石油物探,2005,44(2):163-166.PAN Shulin,GAO Lei,ZOU Qiang,et al.A method for automatic picking of first arrival waves[J].Geophysical Prospecting for Petroleum,2005,44(2):163-166.
[12]
李辉峰,邹强,金文昱.基于边缘检测的初至波自动拾取方法[J].石油地球物理勘探,2006,41(2):150-155.LI Huifeng,ZOU Qiang,JIN Wenyu.Method of automatic first breaks pick-up based on edge detection[J].Oil Geophysical Prospecting,2006,41(2):150-155.
[13]
Boschetti F,Dentith M D,List R D.A fractal-based algorithm for detecting first arrivals onseismic traces[J].Geophysics,1996,61(4):1095-1102.
[14]
左国平.地震记录初至拾取方法对比和研究[D].北京:中国地质大学(北京),2006.ZUO Guoping.The Contrast and Study of First Arrivals Detecting on Seismic Record[D].China Univer-sity of Geosciences (Beijing),Beijing,2006.
兰金涛,张凤军,王春梅,等.基于神经网络算法的可空变初至自动拾取时窗设计[C].中国石油学会2017年物探技术研讨会论文集,2017,197-200.LAN Jintao,ZHANG Fengjun,WANG Chunmei,et al.Design of time window of nullifiable variable first break automatic pickup based on neural network algorithm[C].Proceedings of the 2017 China Geophysical Institute Symposium on Geophysical Prospecting,2017,197-200.
[17]
张玉玺,刘洋,张浩然,等.基于深度学习的多属性盐丘自动识别方法[J].石油地球物理勘探,2020,55(3):475-483.ZHANG Yuxi,LIU Yang,ZHANG Haoran,et al.Multi-attribute salt dome automatic recognition me-thod based on deep learning[J].Oil Geophysical Prospecting,2020,55(3):475-483.
[18]
Yuan S Y,Liu J W,Wang S X,et al.Seismic waveform classification and first-break picking using convolution neural networks[J].IEEE Geoscience and Remote Sensing Letters,2018,15(2):272-276.
[19]
刘佳楠,武杰.基于全卷积神经网络的地震初至波拾取[J].信息技术与网络安全,2018,37(11):62-67.LIU Jianan,WU Jie.Seismic first arrival picking based on fully convolutional networks[J].Information Technology and Network Security,2018,37(11):62-67.
[20]
丁建群,何永清,郭锐,等.基于U-Net网络的地震数据初至自动拾取方法[C].中国地球物理学会2019年油气地球物理学术年会论文集,2019,132-136.DING Jianqun,HE Yongqing,GUO Rui,et al.A method of seismic data first arrival automatic pickup based on U-Net network[C].Proceedings of the 2019 Academic Conference on Petroleum Geophysics,Chinese Geophysical Society,2019,132-136.
[21]
王君,帅威,龚莉,等.基于深度残差网络的初至拾取软件开发[J].石油工业计算机应用,2019,27(1-4):13-16.WANG Jun,SHUAI Wei,GONG Li,et al.Software development of first arrivals pick-up based on deep residual network[J].Computer Applications of Petroleum,2019,27(1-4):13-16.
[22]
帅威,王君,彭延国,等.优化的卷积神经网络初至波自动拾取技术[C].中国地球物理学会2019年油气地球物理学术年会论文集,2019,35-39.SHUAI Wei,WANG Jun,PENG Yanguo,et al.An optimized technique for automatic first arrival pickup based on convolutional neural networks[C].Proceedings of the 2019 Academic Conference on Petroleum Geophysics,Chinese Geophysical Society,2019,35-39.
[23]
浦义涛,张雪利,帅威,等.残差神经网络在地震初至拾取的训练及应用[C].中国石油学会2019年物探技术研讨会电子论文集,2019,1513-1516.FU Yitao,ZHANG Xueli,SHUAI Wei,et al.Training and application of residual neural network in first arrival pickup of seismic[C].Proceedings of the 2019 China Geophysical Institute Symposium on Geophysical Prospecting,Chinese Petroleum Institute,2019,1513-1516.
[24]
张伟,王彦春,李洪臣,等.地震道瞬时强度比法拾取初至波[J].地球物理学进展,2009,24(1):201-204.ZHANG Wei,WANG Yanchun,LI Hongchen,et al.Seismic first arrival pickup using instantaneous intensity ratio[J].Progress in Geophysics,2009,24(1):201-204.
[25]
Ronneberger O,Fischer P,Brox T.U-Net:convolutional networks for biomedical image segmentation[C].Medical Image Computing and Computer Assisted Intervention,2015,234-241.
[26]
Kumar P,Nagar P,Arora C,et al.U-Segnet:fully convolutional neural network based automated brain tissue segmentation tool[C].International Conference on Image Processing,2018,3503-3507.
[27]
Badrinarayanan V,Kendall A,Cipolla R.SegNet:a deep convolutional encoder-decoder architecture for image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(12):2481-2495.
[28]
Ibtehaz N,Rahman M S.MultiResU-Net:rethinking the U-Net architecture for multimodal biomedical image segmentation[J].Neural Networks,2020,doi:10.1016/j.neunet.2019.08.025.
[29]
郭亚萍,顾智聪,彭宏京.SegNet在工件表面缺陷检测中的应用[J].计算机工程与设计,2019,40(10):2979-2984.GUO Yaping,GU Zhicong,PENG Hongjing.Application of SegNet in surface defect detection of workpie-ces[J].Computer Engineering and Design,2019,40(10):2979-2984.
[30]
Keskar N S,Mudigere D,Nocedal J,et al.On large-batch training for deep learning:generalization gap and sharp minima[C].ICLR,2017.
[31]
Ramos D,Francopedroso J,Lozanodiez A,et al.Deconstructing cross-entropy for probabilistic binary classifiers[J].Entropy,2018,doi:10.3390/e20030208.
[32]
郭帅.基于视觉计算的图像目标检测与分析[D].北京:北京邮电大学,2019.GUO Shuai.Image Target Detection and Analysis Based on Visual Computing[D].Beijing University of Posts and Telecommunications,Beijing,2019.
[33]
Kingma D P,Ba J.Adam:a method for stochastic optimization[C].ICLR,2015.