基于Hyperion高光谱数据和随机森林方法的岩性分类与分析

详细信息    查看全文 | 推荐本文 |

英文篇名：Lithological classification and analysis using Hyperion hyperspectral data and Random Forest method 作者：柯元楚 ; 史忠奎 ; 李培军 ; 张西雅 英文作者：KE YuanChu;SHI ZhongKui;LI PeiJun;ZHANG XiYa;Institute of Remote Sensing and GIS,School of Earth and Space Sciences,Peking University;Institute of Urban Meteorology,China Meteorological Administration; 关键词：高光谱遥感数据 ; 随机森林 ; 光谱角制图 ; 岩性分类 英文关键词：Hyperspectral data;;Random Forest;;Spectral angle mapper;;Lithological classification 中文刊名：YSXB 英文刊名：Acta Petrologica Sinica 机构：北京大学地球与空间科学学院遥感与地理信息系统研究所;中国气象局北京城市气象研究所; 出版日期：2018-07-15 出版单位：岩石学报 年：2018 期：v.34 基金：国家自然科学基金项目(41371329)资助 语种：中文; 页：YSXB201807022 页数：8 CN：07 ISSN：11-1922/P 分类号：325-332

摘要

探索利用高光谱数据的岩性填图新方法是遥感地质应用领域的重要需求之一。本文运用随机森林方法和EO-1Hyperion高光谱数据,对新疆塔里木西北部柯坪地区的局部区域进行岩性分类,并对相关问题进行分析。分别利用光谱特征以及加入光谱一阶导数特征进行岩性分类,并对不同特征对岩性分类的重要性进行分析,同时与现有的基于光谱角制图方法(SAM)进行比较。结果表明,与SAM方法相比,随机森林方法得到了更高精度的岩性分类结果,是一种有效可行的岩性分类方法。根据特征重要性的排序,蓝绿光波段、短波红外波段以及相应的一阶导数特征对研究区Hyperion数据的沉积岩岩性分类贡献更大。
        Exploration of new methods for lithological mapping using hyperspectral data is one of critical needs in the field of geological remote sensing applications. In this paper,random forest method was used to classify EO-1 Hyperion hyperspectral data for lithological mapping over a portion of Kalpin,Northwest Tarim,Xinjiang,and the relevant problems were analyzed. The lithological classification was carried out by using spectral features alone and using both spectral features and first derivatives. The importance of different features to lithological classification was calculated and analyzed. The Random Forest method was compared with spectral angle mapper( SAM),a commonly used classification method. The results show that the Random Forest method achieved a significant improvement in accuracy upon lithological classification compared with SAM. Thus,the adopted method is a feasible lithological classification method. According to the ordering of feature importance, blue and green, shortwave infrared bands and their corresponding first derivatives contribute more to lithological classification of sedimentary rocks in the study area.

引文

Amer R,Kusky T and Ghulam A.2010.Lithological mapping in the Central Eastern Desert of Egypt using ASTER data.Journal of African Earth Sciences,56(2):75-82
    Breiman L.1996.Bagging predictors.Machine Learning,24(2):123-140
    Clark RN.1999.Spectroscopy of rocks and minerals,and principles of spectroscopy.In:Rencz AN and Ryerson RA(eds.).Manual of Remote Sensing,Volume 3,Remote Sensing for the Earth Sciences.1999.New York:John Wiley and Sons,3-58
    Cracknell MJ and Reading AM.2014.Geological mapping using remote sensing data:A comparison of five machine learning algorithms,their response to variations in the spatial distribution of training data and the use of explicit spatial information.Computers&Geosciences,63(1):22-33
    Gad S and Kusky T.2006.Lithological mapping in the Eastern Desert of Egypt,the Barramiya area,using Landsat thematic mapper(TM).Journal of African Earth Sciences,44(2):196-202
    Guo YB,Chi TH,Peng L,Liu JL and Yang LN.2016.Classification of GF-1 remote sensing image based on Random Forests for urban landuse.Bulletin of Surveying and Mapping,(5):73-76(in Chinese)
    Ham J,Chen Y,Crawford MM and Ghosh H.2005.Investigation of the Random Forest framework for classification of hyperspectral data.IEEE Transactions on Geoscience&Remote Sensing,43(3):492-501
    Harris JR,Rogge D,Hitchcock R,Ijewliw O and Wright D.2005.Mapping lithology in Canada’s Arctic:Application of hyperspectral data using the minimum noise fraction transformation and matched filtering.Canadian Journal of Earth Sciences,42(12):2173-2193
    Harris JR and Grunsky EC.2015.Predictive lithological mapping of Canada’s North using Random Forest classification applied to geophysical and geochemical data.Computers&Geosciences,80(C):9-25
    Li PJ.2004.Lithological discrimination using ASTER image and geostatistical texture.Journal of Mineralogy and Petrology,24(3):117-121(in Chinese with English abstract)
    Liu M,Lang RL and Cao YB.2015.Number of trees in Random Forest.Computer Engineering and Applications,51(5):126-131(in Chinese with English abstract)
    Masoumi F,Eslamkish T,Abkar AA,Honormand M and Harris JR.2017.Integration of spectral,thermal,and textural features of ASTER data using Random Forests classification for lithological mapping.Journal of African Earth Sciences,129:445-457
    Matthew MW,Adler-Golden SM,Berk A,Richtsmeier SC,Levine RY,Bernstein LS,Acharya PK,Anderson GP,Felde GW,Hoke MP,Ratkowski A,Burke HH,Kaiser RD and Miller DP.2000.Status of atmospheric correction using a MODTRAN4-based algorithm.Proceedings of SPIE-The International Society for Optical Engineering,4049:11
    Meer FDVD,Werff HMAVD,Ruitenbeek FJAV,Hecker CA,Bakker WH,Noomen MF,Meijde MVD,Carranza EJM,Smeth JBD and Woldai T.2012.Multi-and hyperspectral geologic remote sensing:A review.International Journal of Applied Earth Observation&Geoinformation,14(1):112-128
    Murphy RJ,Monteiro ST and Schneider S.2012.Evaluating classification techniques for mapping vertical geology using fieldbased hyperspectral sensors.IEEE Transactions on Geoscience&Remote Sensing,50(8):3066-3080
    Rani N,Mandla VR and Singh T.2016.Performance of image classification on hyperspectral imagery for lithological mapping.Journal of the Geological Society of India,88(4):440-448
    Su HJ and Du PJ.2006.Study on feature selection and extraction of hyperspectral data.Remote Sensing Technology and Application,21(4):288-293(in Chinese with English abstract)
    Xia J,Falco N,Benediktsson JA,Du P and Chanussot J.2017.Hyperspectral image classification with rotation Random Forest Via KPCA.IEEE Journal of Selected Topics in Applied Earth Observations&Remote Sensing,10(4):1601-1609
    Xinjiang Geological Survey.1965.1∶200000 Geological Map of Subashi Area.Beijing:Geological Publishing House(in Chinese)
    Yang JW,Zhang JS,Zhu XF,Xie DF and Yuan ZMQ.2015.Random Forest applied for dimension reduction and classification in hyperspectral data.Journal of Beijing Normal University(Natural Science),51(Suppl.1):82-88
    Zhang XY,Xu HQ and Li PJ.2012.Lithologic mapping using EO-1Hyperion data and extended OCSVM.Acta Scientiarum Naturalium Universitatis,48(3):411-418(in Chinese with English abstract)
    Zhang XY and Li PJ.2014.Lithological mapping from hyperspectral data by improved use of spectral angle mapper.International Journal of Applied Earth Observation&Geoinformation,31(5):95-109
    郭玉宝,池天河,彭玲,刘吉磊,杨丽娜.2016.利用随机森林的高分一号遥感数据进行城市用地分类.测绘通报,(5):73-76
    李培军.2004.用ASETR图像和地统计学纹理进行岩性分类.矿物岩石,24(3):117-121
    刘敏,郎荣玲,曹永斌.2015.随机森林中树的数量.计算机工程与应用,51(5):126-131
    苏红军,杜培军.2006.高光谱数据特征选择与特征提取研究.遥感技术与应用,21(4):288-293
    新疆地质调查局.1965.火焰山地区地质图(1∶20万).北京:地质出版社
    杨珺雯,张锦水,朱秀芳,谢登峰,袁周米奇.2015.随机森林在高光谱遥感数据中降维与分类的应用.北京师范大学学报(自然科学版),51(增1):82-88
    张西雅,徐海卿,李培军.2012.运用EO-1 Hyperion数据和单类支持向量机方法提取岩性信息.北京大学学报(自然科学版),48(3):411-418