多源遥感影像融合及其应用研究
|
摘要
不断进步的遥感技术丰富了可供使用的数据资料。不同类型传感器获取的数据特点不同，多源遥感数据融合技术是综合利用各种数据提取全面地物信息的有效手段。
目前，大多数遥感影像融合研究集中在像素级融合方面，特征级和决策级融合研究相对较少。本文提出了特征级和决策级融合的新方法，并将提出的方法应用于城市“热岛效应”研究和景观格局分析领域。本文主要工作如下：
一、研究多特征融合方法。特征融合的常用方法是串行融合，串行融合只是实现了几何意义上的合并，没有真正有效利用特征之间的相互关系。文中介绍了两种多特征融合方法。本文首次将多视角谱嵌入算法用于遥感影像特征融合。多视角谱嵌入理论利用矩阵等价变换的原理，将不同特征空间的数据变换到同一表示空间，实现不同空间数据的融合。文章提出了基于特征接近度的融合算法。定义吸引力表示特征矢量与原型模式的空间接近程度。用特征矢量与各类型的原型模式之间的吸引力组成接近度矢量，作为融合特征。将这两种算法用于多光谱和高光谱遥感影像地物分类，实验证明，这两种特征融合算法对样本分类正确率超过90%，而且都能有效降低特征矢量维数。
二、研究决策级融合方法。针对多分类器动态集成研究存在的一些问题，如算法的时间复杂度较高，集成过程中需要采用额外的动态机制，动态集成分类的稳定性差，提出了基于判决矩阵的决策融合算法。受行为知识空间融合算法启发，通过训练样本的先验知识提取分类器选择依据。训练样本输入到基本分类器中，得到基本分类器对输入类型的判断。在判决矩阵中记录基本分类器对输入样本类型的判断以及能够对输入正确分类的分类器序号。当输入待分类数据时，根据基本分类器的输出查找判决矩阵，得到样本可以信赖的分类器子集，按照服从多数的原则确定样本类型。分类器的动态选择机制记录在矩阵中，不存在人工参数设定问题，集成结果读取简单。在实验数据和基本分类器相同的情况下，用基于判决矩阵的融合算法和行为知识空间融合算法进行决策融合，基于判决矩阵的融合算法取得了更好的分类效果。
三、以青岛市“热岛效应”研究和青岛市景观格局分析为例，使用本文中提出的方法处理实际问题。在青岛市“热岛效应”研究中发现，人口聚集区域的温度比崂山山区的地面温度高6°左右。研究结果表明，人类的聚集改变了周边环境，人类在活动过程中释放的热能加剧了城市“热岛效应”。景观格局指标显示，青岛市居民区斑块破碎度大，物质流通方便；绿地零星分布在居民区，对缓解城市“热岛效应”作用有限；林地斑块较多，海岸线长，自然条件优越。通过景观格局分析发现，青岛城市化进程对环境影响显著。未来城市发展规划制定应注重环境保护，增强城市宜居指数。
The progress on remote sensing technology enriches available data. Data acquiredby various types of sensors are different. The technology of multi-source remotesensing data fusion is an effective means for the comprehensive utilization of variousdata.
At present, mostly remote sensing imagery fusion research concentrates on thefusion methods of pixel level. Research on feature level and decision level isrelatively little. This paper proposes new fusion methods on feature level and decisionlevel, which are applied in the research of urban "heat island effect" and landscapeanalysis. The main work is as follows.
(I) Multi-feature fusion method is studied. Serial fusion is a common used methodfor feature fusion, which is a combination in geometric meaning and can noteffectively use the correlation between features. In this paper, two algorithms aredescribed to fuse multi-feature. The multi-view spectral embedding is introducedto remote sensing imagery fusion for the first time. Multi-view spectral embeddingtheory is based on the principle of matrix equivalent to transform data fromdifferent feature spaces into the same representation space. Moreover, thepaper presents a novel fusion algorithm based on feature proximity. Attraction isdefined to represent the relationship of feature vector and prototype models. Theproximity vector formed by attraction between prototype models and feature vector isadopted to represent fused feature. The two algorithms are used in multi-spectral andhyper-spectral remote sensing image classification. Experimental results show thatthe two fusion algorithms classify the samples at the correct rate of over90%and they reduce the dimension of feature space significantly.
(II) Decision fusion method is studied. To solve some problems in dynamicintegration of multiple classifiers such as high time complexity, additional dynamicmechanisms, instability of dynamic integration, a decision fusion algorithm based on judgment matrix is proposed. Inspired by the behavior knowledge space fusion, theselection of classifier is based on priori knowledge of the training samples. A trainingsample is inputted into the basic classifiers, and its corresponding type is judged bythe basic classifiers. Record the classification results and the classifier numbers thatcan correctly identify the input type in judgment matrix. In the process of dataclassification, the judgment matrix is accessed according to the output of basicclassifiers to find the reliable classifiers. The type of test sample is obtained inaccordance with the majority. The dynamic selection mechanism is recorded in thematrix so the problem of setting artificial parameters does not exist. It is easy toacquire the integration results. The fusion algorithms based on decision matrix andbehavior knowledge space are performed on the same basic classifiers. Experimentalresults show that the fusion algorithm based on decision matrix can achieve betterclassification results.
(III) The proposed methods are used to analysis the "heat island effect" andlandscape pattern of Qingdao. In the research of "heat island effect", the groundtemperature in densely populated area is6degrees warmer than that in mountainousarea of Laoshan. The results show that the aggregation of human changes thesurrounding environment and energy release in the process of human activitiesaggravates the “heat island effect”. The landscape pattern indexes show thatresidential area patches are fragmentized, which makes material circulationconvenient. The green lands are scattered in residential areas so their ability to relieve"heat island effect" is limited. There are many forest patches and a long coastline,which makes Qingdao own nice natural conditions. The urbanization has a significanteffect on environment. Therefore, the future development plan should pay moreattention to environmental protection and enhance the city livable index.
目前，大多数遥感影像融合研究集中在像素级融合方面，特征级和决策级融合研究相对较少。本文提出了特征级和决策级融合的新方法，并将提出的方法应用于城市“热岛效应”研究和景观格局分析领域。本文主要工作如下：
一、研究多特征融合方法。特征融合的常用方法是串行融合，串行融合只是实现了几何意义上的合并，没有真正有效利用特征之间的相互关系。文中介绍了两种多特征融合方法。本文首次将多视角谱嵌入算法用于遥感影像特征融合。多视角谱嵌入理论利用矩阵等价变换的原理，将不同特征空间的数据变换到同一表示空间，实现不同空间数据的融合。文章提出了基于特征接近度的融合算法。定义吸引力表示特征矢量与原型模式的空间接近程度。用特征矢量与各类型的原型模式之间的吸引力组成接近度矢量，作为融合特征。将这两种算法用于多光谱和高光谱遥感影像地物分类，实验证明，这两种特征融合算法对样本分类正确率超过90%，而且都能有效降低特征矢量维数。
二、研究决策级融合方法。针对多分类器动态集成研究存在的一些问题，如算法的时间复杂度较高，集成过程中需要采用额外的动态机制，动态集成分类的稳定性差，提出了基于判决矩阵的决策融合算法。受行为知识空间融合算法启发，通过训练样本的先验知识提取分类器选择依据。训练样本输入到基本分类器中，得到基本分类器对输入类型的判断。在判决矩阵中记录基本分类器对输入样本类型的判断以及能够对输入正确分类的分类器序号。当输入待分类数据时，根据基本分类器的输出查找判决矩阵，得到样本可以信赖的分类器子集，按照服从多数的原则确定样本类型。分类器的动态选择机制记录在矩阵中，不存在人工参数设定问题，集成结果读取简单。在实验数据和基本分类器相同的情况下，用基于判决矩阵的融合算法和行为知识空间融合算法进行决策融合，基于判决矩阵的融合算法取得了更好的分类效果。
三、以青岛市“热岛效应”研究和青岛市景观格局分析为例，使用本文中提出的方法处理实际问题。在青岛市“热岛效应”研究中发现，人口聚集区域的温度比崂山山区的地面温度高6°左右。研究结果表明，人类的聚集改变了周边环境，人类在活动过程中释放的热能加剧了城市“热岛效应”。景观格局指标显示，青岛市居民区斑块破碎度大，物质流通方便；绿地零星分布在居民区，对缓解城市“热岛效应”作用有限；林地斑块较多，海岸线长，自然条件优越。通过景观格局分析发现，青岛城市化进程对环境影响显著。未来城市发展规划制定应注重环境保护，增强城市宜居指数。
The progress on remote sensing technology enriches available data. Data acquiredby various types of sensors are different. The technology of multi-source remotesensing data fusion is an effective means for the comprehensive utilization of variousdata.
At present, mostly remote sensing imagery fusion research concentrates on thefusion methods of pixel level. Research on feature level and decision level isrelatively little. This paper proposes new fusion methods on feature level and decisionlevel, which are applied in the research of urban "heat island effect" and landscapeanalysis. The main work is as follows.
(I) Multi-feature fusion method is studied. Serial fusion is a common used methodfor feature fusion, which is a combination in geometric meaning and can noteffectively use the correlation between features. In this paper, two algorithms aredescribed to fuse multi-feature. The multi-view spectral embedding is introducedto remote sensing imagery fusion for the first time. Multi-view spectral embeddingtheory is based on the principle of matrix equivalent to transform data fromdifferent feature spaces into the same representation space. Moreover, thepaper presents a novel fusion algorithm based on feature proximity. Attraction isdefined to represent the relationship of feature vector and prototype models. Theproximity vector formed by attraction between prototype models and feature vector isadopted to represent fused feature. The two algorithms are used in multi-spectral andhyper-spectral remote sensing image classification. Experimental results show thatthe two fusion algorithms classify the samples at the correct rate of over90%and they reduce the dimension of feature space significantly.
(II) Decision fusion method is studied. To solve some problems in dynamicintegration of multiple classifiers such as high time complexity, additional dynamicmechanisms, instability of dynamic integration, a decision fusion algorithm based on judgment matrix is proposed. Inspired by the behavior knowledge space fusion, theselection of classifier is based on priori knowledge of the training samples. A trainingsample is inputted into the basic classifiers, and its corresponding type is judged bythe basic classifiers. Record the classification results and the classifier numbers thatcan correctly identify the input type in judgment matrix. In the process of dataclassification, the judgment matrix is accessed according to the output of basicclassifiers to find the reliable classifiers. The type of test sample is obtained inaccordance with the majority. The dynamic selection mechanism is recorded in thematrix so the problem of setting artificial parameters does not exist. It is easy toacquire the integration results. The fusion algorithms based on decision matrix andbehavior knowledge space are performed on the same basic classifiers. Experimentalresults show that the fusion algorithm based on decision matrix can achieve betterclassification results.
(III) The proposed methods are used to analysis the "heat island effect" andlandscape pattern of Qingdao. In the research of "heat island effect", the groundtemperature in densely populated area is6degrees warmer than that in mountainousarea of Laoshan. The results show that the aggregation of human changes thesurrounding environment and energy release in the process of human activitiesaggravates the “heat island effect”. The landscape pattern indexes show thatresidential area patches are fragmentized, which makes material circulationconvenient. The green lands are scattered in residential areas so their ability to relieve"heat island effect" is limited. There are many forest patches and a long coastline,which makes Qingdao own nice natural conditions. The urbanization has a significanteffect on environment. Therefore, the future development plan should pay moreattention to environmental protection and enhance the city livable index.
引文
[1]吴介,裘正定.底层内容特征的融合在图像检索中的研究进展.中国图象图形学报,2008，13（2）：189-197.
[2] El-Melegy M T, Ahmed S M. Neural networks in multiple classifier systems forremote-sensing image classification. Soft Computing in Image Processing, BerlinHeidelberg: Springer.2007.65-94.
[3]征荆,丁晓青,吴佑寿.基于最小代价的多分类器动态集成.计算机学报，1999，22（2）：182-187.
[4]喻莹,杨扬,董才林,等.基于多分类器集成的手写体金融汉字识别.计算机工程与应用,2007,15:212-215.
[5]方敏.集成学习的多分类器动态融合方法研究.系统工程与电子技术,2006,28（11）：1759-1761.
[6]欧吉顺.多分类器动态集成技术研究：[硕士学位论文].镇江：江苏大学,2010.
[7] Haydn, R., G.W. Dalke, and J. Henkel,1982. Application of the IHS transform to theprocessing of multisensor data and image enhancement, Proceedings of InternationalSymposium on Remote Sensing of Arid and Semi-Arid Lands, Cairo, Egypt, pp.599-616.
[8]胡子付，曾志远，张振龙，等.小波和IHS变换结合实现ETM图像波段融合.地球信息科学，2005，7（4）：29-33.
[9]秦鹏,陈健飞.基于Brovey融合与HSV锐化的ASTER影像绿地信息提取分析.地球信息科学学报，2009,11（3）：400-404.
[10] Graham L N. Coincident image fusion using the discrete wavelet transform: a thesis.University of Alabama in Huntsville,1994.
[11]林卉,景海涛,张连蓬. a’Trous小波变换与PCA变换相结合的遥感影像融合分析.地球信息科学，2008,10（2）：269-272.
[12]吴晓荣，何明一，张易凡.基于多小波分解的多光谱图像矢量融合.电子与信息学报，2007，29（4）：789-794.
[13]王攀峰，杜云飞，周海芳，等.基于复小波变换的遥感图像串行融合算法.计算机工程与科学，2008，30（3）：35-39.
[14]于海洋，闫柏琨，甘甫平，等.基于Gram Schmidt变换的高光谱遥感图像改进融合方法.地理与地理信息科学，2007，2（5）：39-42.
[15] Lee D D, Seung H S. Learning the parts of objects by non-negative matrix factorization.Nature,1999,401(6755):788-791.
[16]颜建军，夏春明，郑建荣.基于NMF的多光谱图像和全色图像融合方法.计算机工程，2007,33（21）：169-172.
[17] Kursun O, Alpaydin E, Favorov O V. Canonical correlation analysis using within-classcoupling. Pattern Recognition Letters,2011,32(2):134-144.
[18]杨慧军，敬忠良，赵海涛，等.张量典型相关分析.上海交通大学学报，2008，42（7）：1124-1128.
[19]李弼程，黄洁，高世海，等.信息融合技术及其应用.北京：国防工业出版社，2010.149-172.
[20]宁亚辉,雷小奇,王功孝等. Dempster-Shafer证据融合形状特征的高分辨率遥感图像道路信息提取.中国图象图形学报，2011，16（12）：2183-2190.
[21]汪闽,张星月.多特征证据融合的遥感图像变化检测.遥感学报，2010，14（3）：558-570.
[22] Kauth R, Thomas G. The tasselled cap-A graphical description of the spectral temporaldevelopment of agricultural crops as seen by Landsat. Proceedings of the Symposium onMachine Processing of Remotely Sensed Data. Indiana: Purdue University of WestLafayette,1976，4:41-51.
[23] Chong-gui L I, Xian-wen Z.主分量分析在监测区域样地分布及异常样地探测中的研究.遥感技术与应用，2006，21(1)：1-5.
[24] Gao F, Masek J, Schwaller M, et al. On the blending of the Landsat and MODIS surfacereflectance: Predicting daily Landsat surface reflectance. IEEE Transactions on Geoscienceand Remote Sensing,2006,44:2207-2218.
[25] Chung Y S, Hsu D F, Tang C Y. On the diversity-performance relationship for majorityvoting in classifier ensembles. Multiple Classifier Systems. Berlin Heidelberg: Springer,2007:407-420.
[26] Ceamanos X, Waske B, Benediktsson J A, et al. Ensemble strategies for classifyinghyperspectral remote sensing data. Multiple Classifier Systems. Berlin Heidelberg: Springer,2009:62-71.
[27]邸韡，潘泉，赵永强，等.高光谱图像波段子集模糊积分融合异常检测.电子与信息学报，2008，30（2）：267-271.
[28]马瑶.多尺度几何分析遥感图像融合及分类方法研究：[硕士学位论文].西安：西安电子科技大学，2012.
[29] Wanas N M, Kamel M S. Decision fusion in neural network ensembles. Proceedings ofInternational Joint Conference on Neural Networks. IJCNN: IEEE,2001,4:2952-2957.
[30] Moumen T. El-Melegy，Safaa M. Ahmed. Soft Computing in Image Processing. BerlinHeidelberg: Springer,2007.65-94.
[31]潘竟虎，董晓峰，王建华，等. QuickBird遥感影像融合及在小流域土地利用调查中的应用研究.测绘与空间地理信息，2005，28（6）：13-16.
[32]曹银璇，燕琴，赵争，等. SAR与光学遥感影像融合在土地资源监测中的应用.测绘通报，2007，（8）：23-25.
[33]贺奋琴，胡振琪，尹建忠，等. ASTER和TM/ETM+遥感数据融合监测土地覆盖变化.测绘科学，2007，32（1）：96-97.
[34]尤淑撑，刘顺喜，周连芳，等. CBERS-02B星数据土地利用动态遥感监测方法研究.国土资源遥感，2009，（1）：79-82.
[35]朱运海，张百平，曹银璇，等.土地利用/覆被变化遥感检测方法与应用分析.地球信息科学，2007，9（3）：116-122.
[36]蔡涛，王润生.一个从多波段遥感图像提取道路网的算法.软件学报，2001，12（6）：943-948.
[37]肖志强，鲍光淑，黄继先，等.融合SAR和TM图像更新GIS道路网络数据.测绘学报，2006，35（1）：46-51.
[38]曹广真，金亚秋.基于水平集方法的多源遥感数据融合及城区道路提取.电子与信息学报，2007，29（6）：1464-1470.
[39]千怀遂.农作物遥感估产最佳时相的选择研究--以中国主要粮食作物为例.生态学报，1998，18（1）：48-55.
[40]张霞，张兵，卫征，等. MODIS光谱指数监测小麦长势变化研究.中国图象图形学报，2005，10（4）：420-424.
[41]刘铁梅，曹卫星，罗卫红，等.小麦物质生产与积累的模拟模型.麦类作物学报，2001，21（3）：26-30.
[42]李卫国，王纪华，赵春江，等.基于遥感信息和产量形成过程的小麦估产模型.麦类作物学报，2007，27（5）：904-907.
[43]马艳娟，赵冬玲，王瑞梅，等.基于ASTER数据的近海水产养殖区提取方法对比研究.测绘通报，2011，（1）：59-63.
[44]李俊杰，何隆华，戴锦芳，等.基于遥感影像纹理信息的湖泊围网养殖区提取.湖泊科学，2006，18（4）：337-342.
[45]樊建勇，黄海军，樊辉，等.利用RADARSAT-1数据提取海水养殖区面积.海洋科学，2005，29（10）：44-47.
[46]关学彬，张翠萍，蒋菊生，等.水产养殖遥感监测及信息自动提取方法研究.国土资源遥感，2009，（2）：41-44.
[47]胡兰萍，李燕，张琳，等.遥感FTIR在大气环境监测中的新发展.光谱学与光谱分析，2006，26（10）：1863-1867.
[48]刘军伟，张建民，管海晏，等.卫星数据在延河流域生态环境监测中的应用.中国矿业大学学报，2002，31（5）：394-398.
[49]韩震，金亚秋，恽才兴，等.我国海岸带及其近海资源环境监测的遥感技术应用.遥感信息，2006，（5）：64-66
[50]匡昭敏，陈超泉，黄永璘，等.基于RS和GIS的生态环境监测评估应用系统.计算机工程，2008，34（8）：258-260.
[51]于德浩，龙凡，方洪宾，等.基于数据融合的浅层地下水提取技术研究.国土资源遥感，2010，（3）：114-119.
[52]蒙启安，李国会，苏勇，等.遥感技术在黑龙江省油气勘探中的应用.大庆石油地质与开发，2003，22（2）：16-18.
[53]管海晏，孟淑英，樊新杰，等.遥感找煤深化应用新成果.全国遥感信息综合研究与深化应用交流研讨会论文集，2009：46-48.
[54]丛丽娟，岑况，余学中，等.朱拉扎嘎金矿区地物光谱特征与地球化学成分之间的关系.中南大学学报（自然科学版），2013，44（1）：266-274.
[55] Apan A., Held A., Phinn S., Markley J. Formulation and assessment of narrow-bandvegetation indices from EO-1hyperion imagery for discriminating sugarcane disease, in:Proc.of the Spatial Sciences Conference,2003.22–27.
[56] Everingham Y L, Lowe K H, Donald D A, et al. Advanced satellite imagery to classifysugarcane crop characteristics. Agronomy for sustainable development,2007,27(2):111-117.
[57] Yang C, Everitt J H, Bradford J M, et al. Airborne hyperspectral imagery and yield monitordata for mapping cotton yield variability. Precision Agriculture,2004,5(5):445-461.
[58] Deosthali V, Akmanchi A N. Yield estimation of summer crops grown in a mixed croppingarea, a spectral growth curve profile approach. Journal of the Indian Society of RemoteSensing,2006,34(3):279-288.
[59] Badhwar, G. D. Crop emergence date determination from spectral data. PhotogrammetricEngineering and Remote Sensing,1980,46:369-377.
[60] Johnson, G. V., Raun, W. R.. Nitrogen response index as a guide to fertilizer management.Journal of Plant Nutrition,2003,26:249-262.
[61] Karydas C G, Sekuloska T, Silleos G N. Quantification and site-specification of the supportpractice factor when mapping soil erosion risk associated with olive plantations in theMediterranean island of Crete. Environmental monitoring and assessment,2009,149(1-4):19-28.
[62] Raffaele Casa, Matteo Rossi, Giuseppe Sappa, Antonio Trotta. Assessing crop waterdemand by remote sensing and GIS for the Pontina Plain, Central Italy. Water ResourceManage,2008,4:1-28.
[63] Li Z, Stoll M P, Zhang R, et al. On the separate retrieval of soil and vegetationtemperatures from ATSR data. Science in china series d earth sciences,2001,44(2):97-111.
[64] Franke J, Menz G. Multi-temporal wheat disease detection by multi-spectral remote sensing.Precision Agriculture,2007,8(3):161-172.
[65] Youssef A M. An enhanced remote sensing procedure for material mapping in the westerndesert of Egypt: a tool for managing urban development. Natural resources research,2008,17(4):215-226.
[66] Zeinalov G A. Importance of remote-sensing data in structural geologic analysis of oil-andgas-bearing regions of Azerbaijan. Natural Resources Research,2000,9(4):307-313.
[67] Hochberg E J, Atkinson M J. Coral reef benthic productivity based on optical absorptanceand light-use efficiency. Coral Reefs,2008,27(1):49-59.
[68] Mattio L, Dirberg G, Payri C E, et al. Diversity, biomass and distribution pattern ofSargassum beds in the South West lagoon of New Caledonia (South Pacific). NineteenthInternational Seaweed Symposium. Springer Netherlands,2009.361-373.
[69] Baker M C W. Contribution of landsat imagery to the study of volcanic structures.Geophysical surveys,1982,5(3):281-293.
[70] Carn S A. Application of synthetic aperture radar (SAR) imagery to volcano mapping in thehumid tropics: a case study in East Java, Indonesia. Bulletin of volcanology,1999,61(1-2):92-105.
[71] Jain S K, Singh R D, Jain M K, et al. Delineation of flood-prone areas using remote sensingtechniques. Water Resources Management,2005,19(4):333-347.
[72] King D J, Olthof I, Pellikka P K E, et al. Modelling and mapping damage to forests from anice storm using remote sensing and environmental data. Natural Hazards,2005,35(3):321-342.
[73] Leblon B. Monitoring forest fire danger with remote sensing. Natural Hazards,2005,35(3):343-359.
[74] Lee S, Talib J A. Probabilistic landslide susceptibility and factor effect analysis.Environmental Geology,2005,47(7):982-990.
[75] Sanli F B, Kurucu Y, Esetlili M T. Determining land use changes by radar-optic fusedimages and monitoring its environmental impacts in Edremit region of western Turkey.Environmental monitoring and assessment,2009,151(1-4):45-58.
[76] Coskun H G, Tanik A, Alganci U, et al. Determination of environmental quality of adrinking water reservoir by remote sensing, GIS and regression analysis. Water, Air, andSoil Pollution,2008,194(1-4):275-285.
[77] Taft O W, Haig S M, Kiilsgaard C. Use of radar remote sensing (RADARSAT) to mapwinter wetland habitat for shorebirds in an agricultural landscape. Environmentalmanagement,2003,32(2):268-281.
[78]张春梅，张建明.基于高光谱影像的干旱区草地光谱特征分析.光谱学与光谱分析，2012，32（2）：445-449.
[79]刘娅.模拟蒸发过程中盐渍土反射光谱对水盐变化的响应特征：[硕士学位论文].北京：中国科学院研究生院，2012
[80]张良.多源信息融合方法在铀矿床高光谱数据处理中的应用：[硕士学位论文].武汉：中国地质大学，2010
[81] Rouse, J.W., Haas, R. H., Schell, J. A., Deering, D.W. Monitoring vegetation system ingreat plains with ERTS. Proc.3rd ERTS-1symp.,1973,351:48-62.
[82] Sellers, P. J. Canopy reflectance, photosynthesis, and transpiration. International Journal ofRemote Sensing,1985,6:1335-1372.
[83] Huete, A. R. A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment,1988,25:295-309.
[84]张新乐，张树文，李颖，等.基于光谱角度匹配方法提取黑土边界.光谱学与光谱分析，2009，29（4）：1056-1059.
[85]张玉君，曾朝铭，陈薇，等. ETM+(TM)蚀变遥感异常提取方法研究与应用--方法选择和技术流程.国土资源遥感，2003，（2）：44-49.
[86]张立民，刘峰，张瑞峰，等.一种构造系数的自相关函数特征提取算法.无线电通信技术，2012，38（5）：56-59.
[87]孙小芳，卢健，孙小丹，等.城市地区高分辨率遥感影像绿地提取研究.遥感技术与应用，2006，21（2）：159-162.
[88]张守娟,周诠.卫星遥感图像中的小目标形状描述算法.计算机工程，2011，37（24）：213-215.
[89]瞿继双，瞿松柏，王自杰，等.基于特征的模糊神经网络遥感图像目标分类识别.遥感学报，2009，13（1）：67-74.
[90] Jinmei Liu, Zhongwei Li, Guoyu Wang. Spectrum and spatial invariant based remotesensing image classification. Applied Mechanics and Materials，263-266：905-908.
[91] Haralick R M, Shanmugam K. Texture features for image classification. IEEE Trans. onSys, Man, and Cyb,1973,3(6):610－621.
[92]冯建辉，杨玉静.基于灰度共生矩阵提取纹理特征图像的研究.北京测绘,2007，（3）：19-22.
[93]韦春桃，王宁，张利恒，等.基于纹理特征的高分辨率遥感影像分类方法.桂林理工大学学报，2013，33（1）：80-85.
[94]陈超祥，陈华锋，叶时平，等.高分辨率影像中基于纹理的建筑区信息提取.计算机工程，2011，37（21）：126-130.
[95]陶洪久，柳健，田金文，等.基于小波变换和数学形态学的遥感图像边缘检测.红外与激光工程，2002，31（2）：154-157.
[96]焦斌亮，胡永刚，赵文蕾，等.基于多结构元素的遥感图像去噪及边缘检测方法.宇航学报，2006，27（4）：676-679.
[97]陈高华，卓东风，杨铁梅，等.基于数学形态学的遥感图像边缘检测方法研究.太原科技大学学报，2009，30（4）：280-282.
[98]傅仲良，李勇.基于梯度幅度直方图和类内方差的边缘提取方法.武汉大学学报（信息科学版），2005，30（12）：1056-1058.
[99]崔伟，刘圣霞，徐骞，等.基于互信息和梯度的红外与可见光图像配准新方法.激光与红外，2011，41（2）：224-228.
[100] Long B, Philip S Y, Zhongfei (Mark) Zhang. A general model for multiple viewunsupervised learning. SDM,2008.822-833.
[101]阮秋琦.数字图像处理学.北京：电子工业出版社,2007.423-423
[102] Hunt G R．Electromagnetic radiation：The communication link in remote sensing． NewYork：Wiley，l980：702-702
[103] Rasheed S, Stashuk D. Pattern classification techniques for EMG signal decomposition.Advanced Biosignal Processing. Berlin Heidelberg: Springer2009.267-289.
[104]张涛，洪文学.类空间规整度的计算几何组合分类器权重分配.小型微型计算机系统，2012，33（7）：1572-1576.
[105]陈定三，杨慧中.一种基于改进加权粗糙集的多模型软测量建模方法.化工自动化及仪表，2010，37（1）：11-15.
[106] Cohen J. A coefficient of agreement for nominal scales. Educational and PsychologicalMeasurement,1960,20(2):37–46.
[107] Huang Y S, Suen C Y. A method of combining multiple experts for the recognition ofunconstrained handwritten numerals. Pattern Analysis and Machine Intelligence, IEEETransactions on,1995,17(1):90-94.
[108] Raudys, Roli F. Multiple Classifier Systems. Germany: Springer Berlin Heidelberg,2003.55-64.
[109] Lipnickas A, Korbicz J. Adaptive selection of neural networks for a committee decision.Intelligent Data Acquisition and Advanced Computing Systems: Technology andApplications. Proceedings of the Second IEEE International Workshop on. IEEE,2003.109-114.
[110]窦建奇.关于城市“热岛效应”的思考.武汉城市建设学院学报，2001，18（3）：76-78.
[111]宋治清，王仰麟.城市景观及其格局的生态效应研究进展.地理科学进展，2004，23（2）：97-106.
[112]邬建国.景观生态学——格局、过程、尺度与等级.北京：高等教育出版社，2007.8-26
[113]肖兴媛，张红日，田雨等.青岛市生态环境可持续发展能力动态评价.国土资源科技管理，2009，26（6）：24-28.
[2] El-Melegy M T, Ahmed S M. Neural networks in multiple classifier systems forremote-sensing image classification. Soft Computing in Image Processing, BerlinHeidelberg: Springer.2007.65-94.
[3]征荆,丁晓青,吴佑寿.基于最小代价的多分类器动态集成.计算机学报，1999，22（2）：182-187.
[4]喻莹,杨扬,董才林,等.基于多分类器集成的手写体金融汉字识别.计算机工程与应用,2007,15:212-215.
[5]方敏.集成学习的多分类器动态融合方法研究.系统工程与电子技术,2006,28（11）：1759-1761.
[6]欧吉顺.多分类器动态集成技术研究：[硕士学位论文].镇江：江苏大学,2010.
[7] Haydn, R., G.W. Dalke, and J. Henkel,1982. Application of the IHS transform to theprocessing of multisensor data and image enhancement, Proceedings of InternationalSymposium on Remote Sensing of Arid and Semi-Arid Lands, Cairo, Egypt, pp.599-616.
[8]胡子付，曾志远，张振龙，等.小波和IHS变换结合实现ETM图像波段融合.地球信息科学，2005，7（4）：29-33.
[9]秦鹏,陈健飞.基于Brovey融合与HSV锐化的ASTER影像绿地信息提取分析.地球信息科学学报，2009,11（3）：400-404.
[10] Graham L N. Coincident image fusion using the discrete wavelet transform: a thesis.University of Alabama in Huntsville,1994.
[11]林卉,景海涛,张连蓬. a’Trous小波变换与PCA变换相结合的遥感影像融合分析.地球信息科学，2008,10（2）：269-272.
[12]吴晓荣，何明一，张易凡.基于多小波分解的多光谱图像矢量融合.电子与信息学报，2007，29（4）：789-794.
[13]王攀峰，杜云飞，周海芳，等.基于复小波变换的遥感图像串行融合算法.计算机工程与科学，2008，30（3）：35-39.
[14]于海洋，闫柏琨，甘甫平，等.基于Gram Schmidt变换的高光谱遥感图像改进融合方法.地理与地理信息科学，2007，2（5）：39-42.
[15] Lee D D, Seung H S. Learning the parts of objects by non-negative matrix factorization.Nature,1999,401(6755):788-791.
[16]颜建军，夏春明，郑建荣.基于NMF的多光谱图像和全色图像融合方法.计算机工程，2007,33（21）：169-172.
[17] Kursun O, Alpaydin E, Favorov O V. Canonical correlation analysis using within-classcoupling. Pattern Recognition Letters,2011,32(2):134-144.
[18]杨慧军，敬忠良，赵海涛，等.张量典型相关分析.上海交通大学学报，2008，42（7）：1124-1128.
[19]李弼程，黄洁，高世海，等.信息融合技术及其应用.北京：国防工业出版社，2010.149-172.
[20]宁亚辉,雷小奇,王功孝等. Dempster-Shafer证据融合形状特征的高分辨率遥感图像道路信息提取.中国图象图形学报，2011，16（12）：2183-2190.
[21]汪闽,张星月.多特征证据融合的遥感图像变化检测.遥感学报，2010，14（3）：558-570.
[22] Kauth R, Thomas G. The tasselled cap-A graphical description of the spectral temporaldevelopment of agricultural crops as seen by Landsat. Proceedings of the Symposium onMachine Processing of Remotely Sensed Data. Indiana: Purdue University of WestLafayette,1976，4:41-51.
[23] Chong-gui L I, Xian-wen Z.主分量分析在监测区域样地分布及异常样地探测中的研究.遥感技术与应用，2006，21(1)：1-5.
[24] Gao F, Masek J, Schwaller M, et al. On the blending of the Landsat and MODIS surfacereflectance: Predicting daily Landsat surface reflectance. IEEE Transactions on Geoscienceand Remote Sensing,2006,44:2207-2218.
[25] Chung Y S, Hsu D F, Tang C Y. On the diversity-performance relationship for majorityvoting in classifier ensembles. Multiple Classifier Systems. Berlin Heidelberg: Springer,2007:407-420.
[26] Ceamanos X, Waske B, Benediktsson J A, et al. Ensemble strategies for classifyinghyperspectral remote sensing data. Multiple Classifier Systems. Berlin Heidelberg: Springer,2009:62-71.
[27]邸韡，潘泉，赵永强，等.高光谱图像波段子集模糊积分融合异常检测.电子与信息学报，2008，30（2）：267-271.
[28]马瑶.多尺度几何分析遥感图像融合及分类方法研究：[硕士学位论文].西安：西安电子科技大学，2012.
[29] Wanas N M, Kamel M S. Decision fusion in neural network ensembles. Proceedings ofInternational Joint Conference on Neural Networks. IJCNN: IEEE,2001,4:2952-2957.
[30] Moumen T. El-Melegy，Safaa M. Ahmed. Soft Computing in Image Processing. BerlinHeidelberg: Springer,2007.65-94.
[31]潘竟虎，董晓峰，王建华，等. QuickBird遥感影像融合及在小流域土地利用调查中的应用研究.测绘与空间地理信息，2005，28（6）：13-16.
[32]曹银璇，燕琴，赵争，等. SAR与光学遥感影像融合在土地资源监测中的应用.测绘通报，2007，（8）：23-25.
[33]贺奋琴，胡振琪，尹建忠，等. ASTER和TM/ETM+遥感数据融合监测土地覆盖变化.测绘科学，2007，32（1）：96-97.
[34]尤淑撑，刘顺喜，周连芳，等. CBERS-02B星数据土地利用动态遥感监测方法研究.国土资源遥感，2009，（1）：79-82.
[35]朱运海，张百平，曹银璇，等.土地利用/覆被变化遥感检测方法与应用分析.地球信息科学，2007，9（3）：116-122.
[36]蔡涛，王润生.一个从多波段遥感图像提取道路网的算法.软件学报，2001，12（6）：943-948.
[37]肖志强，鲍光淑，黄继先，等.融合SAR和TM图像更新GIS道路网络数据.测绘学报，2006，35（1）：46-51.
[38]曹广真，金亚秋.基于水平集方法的多源遥感数据融合及城区道路提取.电子与信息学报，2007，29（6）：1464-1470.
[39]千怀遂.农作物遥感估产最佳时相的选择研究--以中国主要粮食作物为例.生态学报，1998，18（1）：48-55.
[40]张霞，张兵，卫征，等. MODIS光谱指数监测小麦长势变化研究.中国图象图形学报，2005，10（4）：420-424.
[41]刘铁梅，曹卫星，罗卫红，等.小麦物质生产与积累的模拟模型.麦类作物学报，2001，21（3）：26-30.
[42]李卫国，王纪华，赵春江，等.基于遥感信息和产量形成过程的小麦估产模型.麦类作物学报，2007，27（5）：904-907.
[43]马艳娟，赵冬玲，王瑞梅，等.基于ASTER数据的近海水产养殖区提取方法对比研究.测绘通报，2011，（1）：59-63.
[44]李俊杰，何隆华，戴锦芳，等.基于遥感影像纹理信息的湖泊围网养殖区提取.湖泊科学，2006，18（4）：337-342.
[45]樊建勇，黄海军，樊辉，等.利用RADARSAT-1数据提取海水养殖区面积.海洋科学，2005，29（10）：44-47.
[46]关学彬，张翠萍，蒋菊生，等.水产养殖遥感监测及信息自动提取方法研究.国土资源遥感，2009，（2）：41-44.
[47]胡兰萍，李燕，张琳，等.遥感FTIR在大气环境监测中的新发展.光谱学与光谱分析，2006，26（10）：1863-1867.
[48]刘军伟，张建民，管海晏，等.卫星数据在延河流域生态环境监测中的应用.中国矿业大学学报，2002，31（5）：394-398.
[49]韩震，金亚秋，恽才兴，等.我国海岸带及其近海资源环境监测的遥感技术应用.遥感信息，2006，（5）：64-66
[50]匡昭敏，陈超泉，黄永璘，等.基于RS和GIS的生态环境监测评估应用系统.计算机工程，2008，34（8）：258-260.
[51]于德浩，龙凡，方洪宾，等.基于数据融合的浅层地下水提取技术研究.国土资源遥感，2010，（3）：114-119.
[52]蒙启安，李国会，苏勇，等.遥感技术在黑龙江省油气勘探中的应用.大庆石油地质与开发，2003，22（2）：16-18.
[53]管海晏，孟淑英，樊新杰，等.遥感找煤深化应用新成果.全国遥感信息综合研究与深化应用交流研讨会论文集，2009：46-48.
[54]丛丽娟，岑况，余学中，等.朱拉扎嘎金矿区地物光谱特征与地球化学成分之间的关系.中南大学学报（自然科学版），2013，44（1）：266-274.
[55] Apan A., Held A., Phinn S., Markley J. Formulation and assessment of narrow-bandvegetation indices from EO-1hyperion imagery for discriminating sugarcane disease, in:Proc.of the Spatial Sciences Conference,2003.22–27.
[56] Everingham Y L, Lowe K H, Donald D A, et al. Advanced satellite imagery to classifysugarcane crop characteristics. Agronomy for sustainable development,2007,27(2):111-117.
[57] Yang C, Everitt J H, Bradford J M, et al. Airborne hyperspectral imagery and yield monitordata for mapping cotton yield variability. Precision Agriculture,2004,5(5):445-461.
[58] Deosthali V, Akmanchi A N. Yield estimation of summer crops grown in a mixed croppingarea, a spectral growth curve profile approach. Journal of the Indian Society of RemoteSensing,2006,34(3):279-288.
[59] Badhwar, G. D. Crop emergence date determination from spectral data. PhotogrammetricEngineering and Remote Sensing,1980,46:369-377.
[60] Johnson, G. V., Raun, W. R.. Nitrogen response index as a guide to fertilizer management.Journal of Plant Nutrition,2003,26:249-262.
[61] Karydas C G, Sekuloska T, Silleos G N. Quantification and site-specification of the supportpractice factor when mapping soil erosion risk associated with olive plantations in theMediterranean island of Crete. Environmental monitoring and assessment,2009,149(1-4):19-28.
[62] Raffaele Casa, Matteo Rossi, Giuseppe Sappa, Antonio Trotta. Assessing crop waterdemand by remote sensing and GIS for the Pontina Plain, Central Italy. Water ResourceManage,2008,4:1-28.
[63] Li Z, Stoll M P, Zhang R, et al. On the separate retrieval of soil and vegetationtemperatures from ATSR data. Science in china series d earth sciences,2001,44(2):97-111.
[64] Franke J, Menz G. Multi-temporal wheat disease detection by multi-spectral remote sensing.Precision Agriculture,2007,8(3):161-172.
[65] Youssef A M. An enhanced remote sensing procedure for material mapping in the westerndesert of Egypt: a tool for managing urban development. Natural resources research,2008,17(4):215-226.
[66] Zeinalov G A. Importance of remote-sensing data in structural geologic analysis of oil-andgas-bearing regions of Azerbaijan. Natural Resources Research,2000,9(4):307-313.
[67] Hochberg E J, Atkinson M J. Coral reef benthic productivity based on optical absorptanceand light-use efficiency. Coral Reefs,2008,27(1):49-59.
[68] Mattio L, Dirberg G, Payri C E, et al. Diversity, biomass and distribution pattern ofSargassum beds in the South West lagoon of New Caledonia (South Pacific). NineteenthInternational Seaweed Symposium. Springer Netherlands,2009.361-373.
[69] Baker M C W. Contribution of landsat imagery to the study of volcanic structures.Geophysical surveys,1982,5(3):281-293.
[70] Carn S A. Application of synthetic aperture radar (SAR) imagery to volcano mapping in thehumid tropics: a case study in East Java, Indonesia. Bulletin of volcanology,1999,61(1-2):92-105.
[71] Jain S K, Singh R D, Jain M K, et al. Delineation of flood-prone areas using remote sensingtechniques. Water Resources Management,2005,19(4):333-347.
[72] King D J, Olthof I, Pellikka P K E, et al. Modelling and mapping damage to forests from anice storm using remote sensing and environmental data. Natural Hazards,2005,35(3):321-342.
[73] Leblon B. Monitoring forest fire danger with remote sensing. Natural Hazards,2005,35(3):343-359.
[74] Lee S, Talib J A. Probabilistic landslide susceptibility and factor effect analysis.Environmental Geology,2005,47(7):982-990.
[75] Sanli F B, Kurucu Y, Esetlili M T. Determining land use changes by radar-optic fusedimages and monitoring its environmental impacts in Edremit region of western Turkey.Environmental monitoring and assessment,2009,151(1-4):45-58.
[76] Coskun H G, Tanik A, Alganci U, et al. Determination of environmental quality of adrinking water reservoir by remote sensing, GIS and regression analysis. Water, Air, andSoil Pollution,2008,194(1-4):275-285.
[77] Taft O W, Haig S M, Kiilsgaard C. Use of radar remote sensing (RADARSAT) to mapwinter wetland habitat for shorebirds in an agricultural landscape. Environmentalmanagement,2003,32(2):268-281.
[78]张春梅，张建明.基于高光谱影像的干旱区草地光谱特征分析.光谱学与光谱分析，2012，32（2）：445-449.
[79]刘娅.模拟蒸发过程中盐渍土反射光谱对水盐变化的响应特征：[硕士学位论文].北京：中国科学院研究生院，2012
[80]张良.多源信息融合方法在铀矿床高光谱数据处理中的应用：[硕士学位论文].武汉：中国地质大学，2010
[81] Rouse, J.W., Haas, R. H., Schell, J. A., Deering, D.W. Monitoring vegetation system ingreat plains with ERTS. Proc.3rd ERTS-1symp.,1973,351:48-62.
[82] Sellers, P. J. Canopy reflectance, photosynthesis, and transpiration. International Journal ofRemote Sensing,1985,6:1335-1372.
[83] Huete, A. R. A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment,1988,25:295-309.
[84]张新乐，张树文，李颖，等.基于光谱角度匹配方法提取黑土边界.光谱学与光谱分析，2009，29（4）：1056-1059.
[85]张玉君，曾朝铭，陈薇，等. ETM+(TM)蚀变遥感异常提取方法研究与应用--方法选择和技术流程.国土资源遥感，2003，（2）：44-49.
[86]张立民，刘峰，张瑞峰，等.一种构造系数的自相关函数特征提取算法.无线电通信技术，2012，38（5）：56-59.
[87]孙小芳，卢健，孙小丹，等.城市地区高分辨率遥感影像绿地提取研究.遥感技术与应用，2006，21（2）：159-162.
[88]张守娟,周诠.卫星遥感图像中的小目标形状描述算法.计算机工程，2011，37（24）：213-215.
[89]瞿继双，瞿松柏，王自杰，等.基于特征的模糊神经网络遥感图像目标分类识别.遥感学报，2009，13（1）：67-74.
[90] Jinmei Liu, Zhongwei Li, Guoyu Wang. Spectrum and spatial invariant based remotesensing image classification. Applied Mechanics and Materials，263-266：905-908.
[91] Haralick R M, Shanmugam K. Texture features for image classification. IEEE Trans. onSys, Man, and Cyb,1973,3(6):610－621.
[92]冯建辉，杨玉静.基于灰度共生矩阵提取纹理特征图像的研究.北京测绘,2007，（3）：19-22.
[93]韦春桃，王宁，张利恒，等.基于纹理特征的高分辨率遥感影像分类方法.桂林理工大学学报，2013，33（1）：80-85.
[94]陈超祥，陈华锋，叶时平，等.高分辨率影像中基于纹理的建筑区信息提取.计算机工程，2011，37（21）：126-130.
[95]陶洪久，柳健，田金文，等.基于小波变换和数学形态学的遥感图像边缘检测.红外与激光工程，2002，31（2）：154-157.
[96]焦斌亮，胡永刚，赵文蕾，等.基于多结构元素的遥感图像去噪及边缘检测方法.宇航学报，2006，27（4）：676-679.
[97]陈高华，卓东风，杨铁梅，等.基于数学形态学的遥感图像边缘检测方法研究.太原科技大学学报，2009，30（4）：280-282.
[98]傅仲良，李勇.基于梯度幅度直方图和类内方差的边缘提取方法.武汉大学学报（信息科学版），2005，30（12）：1056-1058.
[99]崔伟，刘圣霞，徐骞，等.基于互信息和梯度的红外与可见光图像配准新方法.激光与红外，2011，41（2）：224-228.
[100] Long B, Philip S Y, Zhongfei (Mark) Zhang. A general model for multiple viewunsupervised learning. SDM,2008.822-833.
[101]阮秋琦.数字图像处理学.北京：电子工业出版社,2007.423-423
[102] Hunt G R．Electromagnetic radiation：The communication link in remote sensing． NewYork：Wiley，l980：702-702
[103] Rasheed S, Stashuk D. Pattern classification techniques for EMG signal decomposition.Advanced Biosignal Processing. Berlin Heidelberg: Springer2009.267-289.
[104]张涛，洪文学.类空间规整度的计算几何组合分类器权重分配.小型微型计算机系统，2012，33（7）：1572-1576.
[105]陈定三，杨慧中.一种基于改进加权粗糙集的多模型软测量建模方法.化工自动化及仪表，2010，37（1）：11-15.
[106] Cohen J. A coefficient of agreement for nominal scales. Educational and PsychologicalMeasurement,1960,20(2):37–46.
[107] Huang Y S, Suen C Y. A method of combining multiple experts for the recognition ofunconstrained handwritten numerals. Pattern Analysis and Machine Intelligence, IEEETransactions on,1995,17(1):90-94.
[108] Raudys, Roli F. Multiple Classifier Systems. Germany: Springer Berlin Heidelberg,2003.55-64.
[109] Lipnickas A, Korbicz J. Adaptive selection of neural networks for a committee decision.Intelligent Data Acquisition and Advanced Computing Systems: Technology andApplications. Proceedings of the Second IEEE International Workshop on. IEEE,2003.109-114.
[110]窦建奇.关于城市“热岛效应”的思考.武汉城市建设学院学报，2001，18（3）：76-78.
[111]宋治清，王仰麟.城市景观及其格局的生态效应研究进展.地理科学进展，2004，23（2）：97-106.
[112]邬建国.景观生态学——格局、过程、尺度与等级.北京：高等教育出版社，2007.8-26
[113]肖兴媛，张红日，田雨等.青岛市生态环境可持续发展能力动态评价.国土资源科技管理，2009，26（6）：24-28.