一种粒度思想的遥感特征信息识别方法
|
摘要
为了快速分析遥感图像特征信息,结合覆盖分类算法提出了一种商空间理论方法,其中灵活运用商空间理论的分解方法和合成技术来指导遥感图像信息的提取和整合,覆盖算法能快速精确地挖掘出有限信息的本质,为分类提取提供保障.以汶川特大地震后的高级星载热辐射热反射探测仪(ASTER)遥感图像为例进行了实验分析,证实了该方法是快速、精确有效的,可以大大减少遥感图像信息处理的工作量.
Visual interpretation is the main method in remote sensed imagery interpretation,supplemented by computer image processing.It brings some problems,such as strong subjectivity,low accuracy,and high time complexity.In order to analyze feature information of remote sensing image efficiently,a new approach based on the quotient space theory and covering classification algorithm is introduced.It applies decomposition method and synthesis technology of the quotient space theory to extract and integrate remote sensing image in the quotient space theory.Covering classification algorithm is an efficient and constructive neural network algorithm and can wine nature of limited information.By taking Advanced space bone thermal emission and refleetion radiometer(ASTER) remote sensing image of 5.12 Wenchuan earthquakes an example,compared to support vector machine classification algorithm,the result of experiments shows that the method is effective.Therefore,the solution to the problems is undoubtedly a reasonable approach in processing remote sensing image.
Visual interpretation is the main method in remote sensed imagery interpretation,supplemented by computer image processing.It brings some problems,such as strong subjectivity,low accuracy,and high time complexity.In order to analyze feature information of remote sensing image efficiently,a new approach based on the quotient space theory and covering classification algorithm is introduced.It applies decomposition method and synthesis technology of the quotient space theory to extract and integrate remote sensing image in the quotient space theory.Covering classification algorithm is an efficient and constructive neural network algorithm and can wine nature of limited information.By taking Advanced space bone thermal emission and refleetion radiometer(ASTER) remote sensing image of 5.12 Wenchuan earthquakes an example,compared to support vector machine classification algorithm,the result of experiments shows that the method is effective.Therefore,the solution to the problems is undoubtedly a reasonable approach in processing remote sensing image.
引文
[1]Wang Z H,Zhou Y J,Xu B,et al.Preliminaryremote sensing investigation of damage causedby the“5.12”Wenchuan strong earthquake aswell as secondary hazards at the quake center-Yingxiu town.Remote Sensing for Land&Re-sources,2008,2:1~4.(王治华,周英杰,徐斌等.“5.12”汶川大地震震中区映秀镇地震灾情及次生地质灾害遥感初步调查.国土资源遥感,2008,2:1~4).
[2]Wang Z H.Method of estimating volume of thelandslide in the regional landslide remote sens-ing investigation.Proceeding of the 30thInter-national Geological Congress.Geology of FossilFuels Coal,1996.
[3]Zhuo B X.Remote sensing imterpretation andapplication on engineering geology.Beijing:China Railroad Press,2002,1~76.(卓宝熙.工程地质遥感判释与应用.北京:中国铁道出版社,2002,1~76).
[4]Yao X.Fetch and prediction on remote sensinginfor-mation of landslide and debris flow.Ph.DThesis.Beijing:Institute of Geology and Geo-physics,Chinese Academy of Science,2007,23~24.(姚鑫.滑坡泥石流遥感信息获取与预测.博士论文.北京:中国科学院地质与地球物理研究所,2007,23~34).
[5]Yao X,Zhang Y S,Wang X L,et al.Auto-matic hierarchical approach to detecting shallowlandslides and avalanches by the combination ofmulti-spectral RS imagery and DEM deriva-tives.Geological Bulletin of China,2008,27(11):1870~1874.(姚鑫,张永双,王献礼等.基于地貌特征的浅层崩滑体遥感自动识别.地质通报,2008,27(11):1870~1874).
[6]Ye R Q,Deng Q L,Wang H Q.Landsidesidentification based on image classification:Acase study on Guizhou laocheng landslide in thethree gorges reservoir area.Chinese Journal ofEngineering Geophysics,2007,4(6):574~577.(叶润青,邓清禄,王海庆.基于图像分类方法滑坡识别与特征提取———以归州老城滑坡为例.工程地球物理学报,2007,4(6):574~577).
[7]Giles P,Franklin S.An automated approach tothe classi-fication of slope units using digital da-ta.Geomorphology,1998,21:251~264.
[8]Ahmad A R,Khalid M,Yusof R.Kernelmethods and support vector machines for hand-writing recognition.Student Conference on Re-search and Development(SCOReD 2002),ShahAlam,Selangor,2002,309~312.
[9]Bahlmann C,Haasdonk B,Burkhardt H.On-line handwriting recognition with support vectormachines—A kernel approach.Proceedings ofthe 8thInternational Workshop on Frontiers inHandwriting Recognition,Ontario,Canada,2002,49~54.
[10]Wu M,Zhang B,Zhang I.A neural networkbased classifier for handwritten Chinese charac-ter recognition.Proceedings of the 15thInterna-tional Conference on Pattern Recognition,Bar-celona,2000,2:561~568.
[11]Osuna E,Freund R,Girosit F.Training sup-port vector machines:An application to face ce-tection.Proceedings of IEEE Computer SocietyConference on Computer Vision and PatternRecognition,Puerto Rico,1997,130~136.
[12]Buciu L,Kotropoulos C,Pitas I.Combiningsupport vector machines for accurate face detec-tion.Proceedings of International Conferenceon Image Processing,Thessaloniki,2001,1:1054~1057.
[13]Terrillon T J,Shirazi M N.Invariant face de-tection with support vector machines.Proceed-ings of the 15thInternational Conference on Pat-tern Recognition.Barcelona,2000,4:210~217.
[14]Thorsten J.Text categorization with supportvector machines:Learning with many relevantfeatures.Proceedings of the 10thEuropeanConference on machine learning(ECML-98),Springer’Berlin/Heidellerg,1998,137~142.
[15]Cholkopf B,Smola A,Muller K R,et al.Sup-port vector machines in learning and feature ex-traction.Australian Journal of Intelligent Infor-mation Processing Systems,1990,1:3~9.
[16]Zhang L,Zhang B.A geometrical representa-tion of McCulloch-Pitts neural model and its ap-plications.IEEE Transactions on Neural Net-works,1999,10(4):925~929.
[17]Zhang L,Zhang B.A geometrical representa-tion of M-P neural model and its applications.Journal of Software,1998,9(5):334~338.(张铃,张钹.M-P神经元模型的几何意义及其应用.软件学报,1998,9(5):334~338).
[18]Zhang L,Zhang B.The theory and applicationsof problem solving-quotient space based granu-lar computing.The 2ndversion.Beijing:Tsing-hua University Press,2007,1~188.(张铃,张钹.问题求解理论及应用.第2版.北京:清华大学出版社,2007,1~188).
[19]Zhang L,Zhang B,Yin H F.An alternativecovering design algorithm of multi-layer neuralnetworks.Journal of Software,1999,10(7):737~742.(张铃,张钹,殷海风.多层前向网络的交叉覆盖设计算法.软件学报,1999,7:737~742).
[2]Wang Z H.Method of estimating volume of thelandslide in the regional landslide remote sens-ing investigation.Proceeding of the 30thInter-national Geological Congress.Geology of FossilFuels Coal,1996.
[3]Zhuo B X.Remote sensing imterpretation andapplication on engineering geology.Beijing:China Railroad Press,2002,1~76.(卓宝熙.工程地质遥感判释与应用.北京:中国铁道出版社,2002,1~76).
[4]Yao X.Fetch and prediction on remote sensinginfor-mation of landslide and debris flow.Ph.DThesis.Beijing:Institute of Geology and Geo-physics,Chinese Academy of Science,2007,23~24.(姚鑫.滑坡泥石流遥感信息获取与预测.博士论文.北京:中国科学院地质与地球物理研究所,2007,23~34).
[5]Yao X,Zhang Y S,Wang X L,et al.Auto-matic hierarchical approach to detecting shallowlandslides and avalanches by the combination ofmulti-spectral RS imagery and DEM deriva-tives.Geological Bulletin of China,2008,27(11):1870~1874.(姚鑫,张永双,王献礼等.基于地貌特征的浅层崩滑体遥感自动识别.地质通报,2008,27(11):1870~1874).
[6]Ye R Q,Deng Q L,Wang H Q.Landsidesidentification based on image classification:Acase study on Guizhou laocheng landslide in thethree gorges reservoir area.Chinese Journal ofEngineering Geophysics,2007,4(6):574~577.(叶润青,邓清禄,王海庆.基于图像分类方法滑坡识别与特征提取———以归州老城滑坡为例.工程地球物理学报,2007,4(6):574~577).
[7]Giles P,Franklin S.An automated approach tothe classi-fication of slope units using digital da-ta.Geomorphology,1998,21:251~264.
[8]Ahmad A R,Khalid M,Yusof R.Kernelmethods and support vector machines for hand-writing recognition.Student Conference on Re-search and Development(SCOReD 2002),ShahAlam,Selangor,2002,309~312.
[9]Bahlmann C,Haasdonk B,Burkhardt H.On-line handwriting recognition with support vectormachines—A kernel approach.Proceedings ofthe 8thInternational Workshop on Frontiers inHandwriting Recognition,Ontario,Canada,2002,49~54.
[10]Wu M,Zhang B,Zhang I.A neural networkbased classifier for handwritten Chinese charac-ter recognition.Proceedings of the 15thInterna-tional Conference on Pattern Recognition,Bar-celona,2000,2:561~568.
[11]Osuna E,Freund R,Girosit F.Training sup-port vector machines:An application to face ce-tection.Proceedings of IEEE Computer SocietyConference on Computer Vision and PatternRecognition,Puerto Rico,1997,130~136.
[12]Buciu L,Kotropoulos C,Pitas I.Combiningsupport vector machines for accurate face detec-tion.Proceedings of International Conferenceon Image Processing,Thessaloniki,2001,1:1054~1057.
[13]Terrillon T J,Shirazi M N.Invariant face de-tection with support vector machines.Proceed-ings of the 15thInternational Conference on Pat-tern Recognition.Barcelona,2000,4:210~217.
[14]Thorsten J.Text categorization with supportvector machines:Learning with many relevantfeatures.Proceedings of the 10thEuropeanConference on machine learning(ECML-98),Springer’Berlin/Heidellerg,1998,137~142.
[15]Cholkopf B,Smola A,Muller K R,et al.Sup-port vector machines in learning and feature ex-traction.Australian Journal of Intelligent Infor-mation Processing Systems,1990,1:3~9.
[16]Zhang L,Zhang B.A geometrical representa-tion of McCulloch-Pitts neural model and its ap-plications.IEEE Transactions on Neural Net-works,1999,10(4):925~929.
[17]Zhang L,Zhang B.A geometrical representa-tion of M-P neural model and its applications.Journal of Software,1998,9(5):334~338.(张铃,张钹.M-P神经元模型的几何意义及其应用.软件学报,1998,9(5):334~338).
[18]Zhang L,Zhang B.The theory and applicationsof problem solving-quotient space based granu-lar computing.The 2ndversion.Beijing:Tsing-hua University Press,2007,1~188.(张铃,张钹.问题求解理论及应用.第2版.北京:清华大学出版社,2007,1~188).
[19]Zhang L,Zhang B,Yin H F.An alternativecovering design algorithm of multi-layer neuralnetworks.Journal of Software,1999,10(7):737~742.(张铃,张钹,殷海风.多层前向网络的交叉覆盖设计算法.软件学报,1999,7:737~742).
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心