遥感数据可靠性分类方法研究
|
摘要
遥感数据分类技术是从遥感数据中提取专题类别数据的一个重要手段，但由于自然环境的复杂性、遥感传感器及分类算法的局限性等原因，不确定性伴随遥感数据分类的整个过程，在利用遥感数据分类获取的信息中含有较大的不确定性。如何了解这些不确定性的本质，降低其对分类结果精度的影响，提高分类精度，建立可靠的遥感数据分类方法，是遥感数据分类研究的重要问题。为此，本文以遥感数据分类过程为出发点，研究分类过程中的主要环节的不确定性对分类精度的影响，提出遥感数据可靠性分类方法，提高最终分类结果精度。研究内容包括训练样本数据对遥感数据分类精度的影响、遥感数据可靠性分类模型和分类精度评价的可靠性样本抽样方法三个部分。研究成果将为提高遥感数据分类结果的精度提供一整套新的解决方案。具体研究工作主要包括：
（1）通过对典型区域遥感数据分类实验的研究，得出训练样本数量、质量及抽样方法与不同分类方法分类精度的响应关系：1）不同的分类方法对样本量的响应及同种分类方法对相同样本量的响应程度是不同的，且分类精度都存在一定程度的波动性，但达到一定数据的样本时，分类精度均值是相对稳定的；2）同一种分类方法利用在相同质量指标划分的不同质量等级的训练样本得到的分类精度不同；不同分类方法对在相同指标划分的同一质量等级的训练样本响应也是不同的；3）不同抽样方式下的多次抽样获取的训练样本得到的分类结果精度的平均值基本上都能够反映实际的分类精度，分层抽样方法要比非分层抽样方法好，点抽样方法要比群抽样方法好。
（2）针对混合像元是造成分类精度低的根本原因，将模糊拓扑理论引入到传统SVM，提出比传统SVM分类精度更高的FTSVM遥感影像分类方法。该方法利用提出的改进投票法得到每个像元的后验概率，然后根据最优阈值确定每个像元属于类别的内部、边界或外部，最后利用模糊拓扑中的空间邻域相关性对处于类别边界的像元进行重新分类，从而提高了混合像元的分类精度，最终提高了整个遥感数据分类精度。
（3）充分考虑不同分类器对每个像元的识别能力，提出了基于矩阵特征向量的多分类器组合方法，利用多个分类器的输出的每个像元的概率矢量构成一个概率矢量矩阵，然后根据每个像元的概率矩阵的特性，自适应调整每个分类器对此像元的权值，为分类性能好（分类不确定性小）的分类器赋予较大的权重，提高了分类精度和稳定性。
（4）提出融合光谱与空间特征的遥感数据多种分类方法：1）充分考虑像元之间的邻域信息，提出改进后的FCM图像分割模型MFCM，获取图像的同质区域并与基于像元的MLC分类结果采用投票法融合，获得了比传统的MLC方法分类精度高的分类结果图；2）根据地理学第一定理的邻近原则，离中心像元距离越远的像元的影响应越小，引入空间引力模型使用距离函数表达像元间的相互作用在距离上的非线性关系，对经典的MRF模型进行改进，提出了SAMRF模型，使得其更符合实际情况，增强了对类别边缘的描述能力，提高了分类精度和稳定性；3）在分类中融合了Gabor小波、GMRF和GLCM三种纹理特征，并提出了PAI、LW、Solidity和Extent四个像元形状特征，通过实验验证了纹理、形状特征及纹理融合、形状特征融合以及这两大类特征组合可以大大提高分类精度，且不同的纹理特征及形状特征对于不同的遥感影像中不同地类的分类精度的提高程度不同，在实际分类过程中，应将多个纹理和像元形状特征进行组合用于遥感影像分类，对于多特征组合分类，一般先要进行特征组合优化。
（5）提出了基于空间均衡抽样和基于聚类的空间分层抽样方法进行检验样本选择，保证了验证样本的均匀性和代表性，从而提高了精度评价结果的可靠性。
Remotely sensed image classification is an important technique for extractingthematic mapping information. Due to the complexity of interactions in the naturalenvironment as well as the natural environment and remote sensing spectroscopy,surface information extracted from the sensor spectral signal includes uncertainties.Many uncertainties are inevitably introduced and spread during the process ofclassification. The nature of uncertainty must be known when using these dataincluding uncertainty, reducing its impact on the classification accuracy and obtainingreliable classification methods to improve the classification accuracy. The objective ofstudy is to provide reliable remotely sensed image classification methods. The processof remotely sensed image classification being as the starting point, uncertaintiesimpacting on the classification accuracy are studied in the main parts of theclassification processing, and reliable classification methods are presented to reducethe uncertainties in the results so as to improve the final classification accuracy. Thestudy has important theoretical significance and application value in extractinginformation from the remotely sensed image, remote sensing applications and spatialdata uncertainty. The body mainly includes effect of training samples on remotelysensed image classification accuracy, reliable remotely sensed image classificationmethods and reliable sampling methods for accuracy assessment, etc. The study willprovide a new method for improving the remotely sensed image classificationaccuracy. The details are as follows:
(1) Relationships between quantity, quality and sampling method of trainingsamples and classification accuracies were concluded from experiments are as follows:1) Different classification methods’ responses to different size of samples areinequable, as well as the same method’s response to the same size of samples, thefinal classification accuracies trendings show a degree of volatilities, but the mean ofaccuracies is relatively stable when the size of samples is above certain size.2) Thequality of samples had major Effect on the classification accuracy, the accuracies aredifferent using different classification methods based on different quality of samplesunder different quality criterions. Different classification methods’ responses todifferent quality of sample under different quality criterions are roughly the same.Different classification methods’ responses to the same quality of samples under thesame quality criterions are different.3) Remotely sensed image classificationaccuracies are inevitable different based on training samples using different sampling methods, but the mean of classification accuracies can describe the true classificationresult based on training samples using different sampling methods. Stratified samplingmethod is better the non-stratified sampling method, and point sampling method isbetter than cluster sampling method.
(2) Considering the mixed pixels being as the major factor resulting in the lowclassification accuracy and most of mixed pixels distribute on the boundary ofdifferent classes, the dissertation introduces fuzzy topology theory and presents anovel fuzzy-topology integrated support vector machine (SVM)(FTSVM)classification method for remotely sensed images based on the standard SVM. First,the optimal inter-correlation coefficient threshold value is applied to decompose animage class in spectral space into the three parts: interior, boundary, and exterior infuzzy-topology space. The interior class pixels are then classified as predefinedclasses based on maximum likelihood. The exterior-class pixels are ignored. Thefuzzy-boundary-class pixels which contain misclassified pixels are reclassified basedon the fuzzy-topology connectivity theory. As a result, misclassified pixel problems,to a certain extent, are solved, and the classification accuracy is improved.
(3) The dissertation presents a new eigen-values-based multiple classifierscombination, in which, considering the classification capacity for every pixels basedon different classifiers, first, a probability vector matrix for every pixel is obtainedbased on the probability vector output from each classifier, then, the differentclassifiers’ weights on a pixel are decided based on the characteristics of theprobability vector matrix, a bigger weight will be attached on the good performanceclassifier, as a result, the classification accuracy and stability are improved.
(4) The dissertation presents many remotely sensed image classification methodsby fusion of multiple spatial features. The details are as follows:
1) Presenting a remotely sensed image classification method based on MarkovRandom Field-based Fuzzy C-means Clustering Algorithm (MFCM), the proposedmethod combines the results of a pixel wise spectral classification and a segmentationmap, aiming to improve classification accuracy, when compared to pixel wiseclassification only, in which the neighborhood information of pixel are fullyconsidered, first, the homogeneous regions are extracted using MFCM from theremotely sensed image, thereby, the spatial contextual information of pixels can beobtained from the homogeneous regions, then, the final classification result map isobtained by combining homogeneous regions map and the results of a pixel wise spectral classification and a segmentation map by majority voting.
2) Presenting a remotely sensed image classification method based on SpatialAttraction-based Markov Random Field (SAMRF), pixels’impacts on the center pixelare inversely proportion to their distances from the center pixel under the Tbler's FirstLaw of Geography, and spatial attraction (SA) model is introduced to express thenon-linear relationship at a distance about the interaction among the pixels, theEqual-weighted MRF (EWMRF) is improved to accord with the fact.
3) Presenting many remotely sensed image classification methods by fusionspectral, textures and pixel shape features. The remotely sensed image is classified bydifferent fusions of Gabor wavelet, Gaussian Markov Random Fields (GMRF) andGrey-Level Co-occurrence Matrix (GLCM) textures and proposed PAI, LW, Solidityand Extent pixel shape features to improve the classification accuracy. Experimentalresults indicate that classification accuracy can be greatly improved by fusions oftextures, pixel shape indices or textures and pixel shape features, and improvementsof every class in the image are different by different fusions of textures or pixel shapefeatures. Generally, GLCM performs better than the other textures. Alone pixel shapefeature cannot improve every class’s classification accuracy, thus, multiple pixelshape features should be fused for remotely sensed image classification to obtainsatisfactory classification accuracy. Feature Selection always is performed in terms ofclassification by fusion of multiple features.
(5) The dissertation presents Spatially Balanced Sampling (SBS) andCluster-based spatial stratified sampling for obtaining uniform and representativetesting samples, thus, the reliability of accuracy assessment result is improved.
（1）通过对典型区域遥感数据分类实验的研究，得出训练样本数量、质量及抽样方法与不同分类方法分类精度的响应关系：1）不同的分类方法对样本量的响应及同种分类方法对相同样本量的响应程度是不同的，且分类精度都存在一定程度的波动性，但达到一定数据的样本时，分类精度均值是相对稳定的；2）同一种分类方法利用在相同质量指标划分的不同质量等级的训练样本得到的分类精度不同；不同分类方法对在相同指标划分的同一质量等级的训练样本响应也是不同的；3）不同抽样方式下的多次抽样获取的训练样本得到的分类结果精度的平均值基本上都能够反映实际的分类精度，分层抽样方法要比非分层抽样方法好，点抽样方法要比群抽样方法好。
（2）针对混合像元是造成分类精度低的根本原因，将模糊拓扑理论引入到传统SVM，提出比传统SVM分类精度更高的FTSVM遥感影像分类方法。该方法利用提出的改进投票法得到每个像元的后验概率，然后根据最优阈值确定每个像元属于类别的内部、边界或外部，最后利用模糊拓扑中的空间邻域相关性对处于类别边界的像元进行重新分类，从而提高了混合像元的分类精度，最终提高了整个遥感数据分类精度。
（3）充分考虑不同分类器对每个像元的识别能力，提出了基于矩阵特征向量的多分类器组合方法，利用多个分类器的输出的每个像元的概率矢量构成一个概率矢量矩阵，然后根据每个像元的概率矩阵的特性，自适应调整每个分类器对此像元的权值，为分类性能好（分类不确定性小）的分类器赋予较大的权重，提高了分类精度和稳定性。
（4）提出融合光谱与空间特征的遥感数据多种分类方法：1）充分考虑像元之间的邻域信息，提出改进后的FCM图像分割模型MFCM，获取图像的同质区域并与基于像元的MLC分类结果采用投票法融合，获得了比传统的MLC方法分类精度高的分类结果图；2）根据地理学第一定理的邻近原则，离中心像元距离越远的像元的影响应越小，引入空间引力模型使用距离函数表达像元间的相互作用在距离上的非线性关系，对经典的MRF模型进行改进，提出了SAMRF模型，使得其更符合实际情况，增强了对类别边缘的描述能力，提高了分类精度和稳定性；3）在分类中融合了Gabor小波、GMRF和GLCM三种纹理特征，并提出了PAI、LW、Solidity和Extent四个像元形状特征，通过实验验证了纹理、形状特征及纹理融合、形状特征融合以及这两大类特征组合可以大大提高分类精度，且不同的纹理特征及形状特征对于不同的遥感影像中不同地类的分类精度的提高程度不同，在实际分类过程中，应将多个纹理和像元形状特征进行组合用于遥感影像分类，对于多特征组合分类，一般先要进行特征组合优化。
（5）提出了基于空间均衡抽样和基于聚类的空间分层抽样方法进行检验样本选择，保证了验证样本的均匀性和代表性，从而提高了精度评价结果的可靠性。
Remotely sensed image classification is an important technique for extractingthematic mapping information. Due to the complexity of interactions in the naturalenvironment as well as the natural environment and remote sensing spectroscopy,surface information extracted from the sensor spectral signal includes uncertainties.Many uncertainties are inevitably introduced and spread during the process ofclassification. The nature of uncertainty must be known when using these dataincluding uncertainty, reducing its impact on the classification accuracy and obtainingreliable classification methods to improve the classification accuracy. The objective ofstudy is to provide reliable remotely sensed image classification methods. The processof remotely sensed image classification being as the starting point, uncertaintiesimpacting on the classification accuracy are studied in the main parts of theclassification processing, and reliable classification methods are presented to reducethe uncertainties in the results so as to improve the final classification accuracy. Thestudy has important theoretical significance and application value in extractinginformation from the remotely sensed image, remote sensing applications and spatialdata uncertainty. The body mainly includes effect of training samples on remotelysensed image classification accuracy, reliable remotely sensed image classificationmethods and reliable sampling methods for accuracy assessment, etc. The study willprovide a new method for improving the remotely sensed image classificationaccuracy. The details are as follows:
(1) Relationships between quantity, quality and sampling method of trainingsamples and classification accuracies were concluded from experiments are as follows:1) Different classification methods’ responses to different size of samples areinequable, as well as the same method’s response to the same size of samples, thefinal classification accuracies trendings show a degree of volatilities, but the mean ofaccuracies is relatively stable when the size of samples is above certain size.2) Thequality of samples had major Effect on the classification accuracy, the accuracies aredifferent using different classification methods based on different quality of samplesunder different quality criterions. Different classification methods’ responses todifferent quality of sample under different quality criterions are roughly the same.Different classification methods’ responses to the same quality of samples under thesame quality criterions are different.3) Remotely sensed image classificationaccuracies are inevitable different based on training samples using different sampling methods, but the mean of classification accuracies can describe the true classificationresult based on training samples using different sampling methods. Stratified samplingmethod is better the non-stratified sampling method, and point sampling method isbetter than cluster sampling method.
(2) Considering the mixed pixels being as the major factor resulting in the lowclassification accuracy and most of mixed pixels distribute on the boundary ofdifferent classes, the dissertation introduces fuzzy topology theory and presents anovel fuzzy-topology integrated support vector machine (SVM)(FTSVM)classification method for remotely sensed images based on the standard SVM. First,the optimal inter-correlation coefficient threshold value is applied to decompose animage class in spectral space into the three parts: interior, boundary, and exterior infuzzy-topology space. The interior class pixels are then classified as predefinedclasses based on maximum likelihood. The exterior-class pixels are ignored. Thefuzzy-boundary-class pixels which contain misclassified pixels are reclassified basedon the fuzzy-topology connectivity theory. As a result, misclassified pixel problems,to a certain extent, are solved, and the classification accuracy is improved.
(3) The dissertation presents a new eigen-values-based multiple classifierscombination, in which, considering the classification capacity for every pixels basedon different classifiers, first, a probability vector matrix for every pixel is obtainedbased on the probability vector output from each classifier, then, the differentclassifiers’ weights on a pixel are decided based on the characteristics of theprobability vector matrix, a bigger weight will be attached on the good performanceclassifier, as a result, the classification accuracy and stability are improved.
(4) The dissertation presents many remotely sensed image classification methodsby fusion of multiple spatial features. The details are as follows:
1) Presenting a remotely sensed image classification method based on MarkovRandom Field-based Fuzzy C-means Clustering Algorithm (MFCM), the proposedmethod combines the results of a pixel wise spectral classification and a segmentationmap, aiming to improve classification accuracy, when compared to pixel wiseclassification only, in which the neighborhood information of pixel are fullyconsidered, first, the homogeneous regions are extracted using MFCM from theremotely sensed image, thereby, the spatial contextual information of pixels can beobtained from the homogeneous regions, then, the final classification result map isobtained by combining homogeneous regions map and the results of a pixel wise spectral classification and a segmentation map by majority voting.
2) Presenting a remotely sensed image classification method based on SpatialAttraction-based Markov Random Field (SAMRF), pixels’impacts on the center pixelare inversely proportion to their distances from the center pixel under the Tbler's FirstLaw of Geography, and spatial attraction (SA) model is introduced to express thenon-linear relationship at a distance about the interaction among the pixels, theEqual-weighted MRF (EWMRF) is improved to accord with the fact.
3) Presenting many remotely sensed image classification methods by fusionspectral, textures and pixel shape features. The remotely sensed image is classified bydifferent fusions of Gabor wavelet, Gaussian Markov Random Fields (GMRF) andGrey-Level Co-occurrence Matrix (GLCM) textures and proposed PAI, LW, Solidityand Extent pixel shape features to improve the classification accuracy. Experimentalresults indicate that classification accuracy can be greatly improved by fusions oftextures, pixel shape indices or textures and pixel shape features, and improvementsof every class in the image are different by different fusions of textures or pixel shapefeatures. Generally, GLCM performs better than the other textures. Alone pixel shapefeature cannot improve every class’s classification accuracy, thus, multiple pixelshape features should be fused for remotely sensed image classification to obtainsatisfactory classification accuracy. Feature Selection always is performed in terms ofclassification by fusion of multiple features.
(5) The dissertation presents Spatially Balanced Sampling (SBS) andCluster-based spatial stratified sampling for obtaining uniform and representativetesting samples, thus, the reliability of accuracy assessment result is improved.
引文
[1] Lunetta R, Congalton R G and Fenstermaker L K. Remote sensing and geographic informationsystem data integration: error sources and research issues[J]. Photogrammetric Engineeringand Remote Sensing,1991,57(6):677-687.
[2]史文中,陈江平,詹庆明等.可靠性空间分析初探[J].武汉大学学报(信息科学版),2012,37(8):883-991.
[3] Congalton R G. A review of assessing the accuracy of classifications of remotely senseddata[J]. Remote Sensing of Environment,1991,37(1):35-46.
[4] Shi W Z. Principle of modeling uncertainties in spatial data and spatial analyses[M]. New York:Taylor&Francis Group/CRC Press,2009:85.
[5](美)延森(Jensen, J R)著;陈晓玲等译.遥感数字影像处理导论(原书第3版)[M].北京:机械工业出版社,2007:325-326.
[6] Han J and Kamber M. Data mining: concepts and techniques[M]. Los Altos, CA, USA:Morgan Kaufmann Publishers,2001:142.
[7] Hixson M, Scholz D, Fuhs N and Akiyama T. Evaluation of several schemes for classificationof remotely sensed data[J]. Photogrammetric Engineering and Remote Sensing,1990,46(1)：1547-1553.
[8] Jackson Q and Landgrebe D A. An adaptive classifier design for high-dimensional dataanalysis with a limited training data set[J]. IEEE Transactions on Geoscience and RemoteSensing,2001,39(12):2664-2679.
[9] Nelson R F, Latty R S and Mott G. Classifying northern forests using Thematic Mappersimulator data[J]. Photogrammetric Engineering and Remote Sensing,1984,50(5):607-617.
[10] Van Niel T G, McVicar T R and Datt B. On the relationship between training sample size anddata dimensionality of broadband multi-temporal classification[J]. Remote Sensing ofEnvironment,2005,98(4):468-480.
[11] Zhuang X, Engel B A and Lozano-Garcia D F, et al. Optimisation of training data required forneuro-classification[J]. International Journal of Remote Sensing,1994,15(16):3271-3277.
[12] Foody G M and Arora M K. An evaluation of some factors affecting the accuracy ofclassification by an artificial neural network[J]. International Journal of Remote Sensing,1997,18(4):799-810.
[13]阎静,王汶,李湘阁.利用神经网络方法提取水稻种植面积-以湖北省双季早稻为例[J].遥感学报,2001,5(3):227-231.
[14] Foody G M, Mathur A. Toward intelligent training of supervised image classifications:directing training data acquisition for SVM classification[J]. Remote Sensing of Environment,2004,93(1-2):107-117.
[15] Foody G M. Thematic map comparison: Evaluating the statistical significance of differencesin classification accuracy[J]. Photogrammetric Engineering and Remote Sensing,2004,70(5):627-633.
[16] Foody G M, Mathur A, Sanchez-Hernandez C and Boyd D S. Training set size requirementsfor the classification of a specific class[J]. Remote Sensing of Environment,2006,104(1):1-14.
[17]朱秀芳,潘耀忠,张锦水等.训练样本对TM尺度小麦种植面积测量精度影响研究(Ⅰ)-训练样本与分类方法间分类精度响应关系研究[J].遥感学报,2007,11(6):826-837.
[18]薄树奎,丁琳.训练样本数目选择对面向对象影像分类方法精度的影响[J].中国图象图形学报,2010,15(7):1106-1111.
[19] Kuo B C and Landgrebe D A. A covariance estimator for small sample size classificationproblems and its application to feature extraction[J]. IEEE Transactions on Geoscience andRemote Sensing,2002,40(4):814-819.
[20] Arai K. A supervised thematic mapper classification with a purification of training samples.International Journal of Remote Sensing,1992,13(11):2039-2049.
[21]吴健平,杨星卫.遥感数据监督分类中训练样本的纯化[J].国土资源遥感,1996,27(1):36-41.
[22] Foody G M and Mathur A. The use of small training sets containing mixed pixels foraccurate hard image classification: Training on mixed spectral responses for classification bya SVM [J]. Remote Sensing of Environment,2006,103,(2),179-189.
[23] Schmidt K S and Skidmore A K. Spectral discrimination of vegetation types in a coastalwetland[J]. Remote Sensing of Environment,2003,85(1):92-108.
[24]杜红艳,张洪岩,张正祥. GIS支持下的湿地遥感信息高精度分类方法研究[J].遥感技术与应用,2004,19(4):244-248.
[25]连石柱.基于分离度的图像特征提取于识别[J].中国图象图形学报,1996,1(3):196-200.
[26]张治英,徐德忠,周云等.不同特征遥感图像在江滩钉螺孳生地监测中的应用[J].西安交通大学学报,2004,25(3):304-307.
[27] Key T, Warner T A and McGraw J B, et al. A comparison of multispectral and multitemporalinformation in high spatial resolution imagery for classification of individual tree species in atemperate hardwood forest[J]. Remote Sensing of Environment,2001,75(1):100-112.
[28] Koukoulas S and George Alan B. Introducing new indices for accuracy evaluation ofclassified images representing semi-natural woodland environments[J]. PhotogrammetricEngineering and Remote Sensing,2001,67(4):499-510.
[29] Van Niela T G, McVicarb T R and Dattc B. On the relationship between training sample sizeand data dimensionality: Monte Carlo analysis of broadband multi-temporal classification[J].Remote Sensing of Environment,2005,98(4):468-480.
[30] Scott. C T. Sampling method for estimating change in forest resources[J]. EcologicalApplications,1998,8(2):228-233.
[31] Dunn R and Harrison A R. Two-dimensional systematic sampling of land use[J]. AppliedStatistics,1993,42(4):585-601.
[32]曹志冬,王劲峰,李连发等.地理空间中不同分层抽样方式的分层效率与优化策略[J].地理科学进展,2008,27(3):152-160.
[33]冯士雍.抽样调查应用与理论中的若干前沿问题[J].统计与信息论坛,2007,22(1):5-31.
[34]冯士雍,施锡铨.抽样调查:理论、方法与实践[M].上海:上海科学技术出版社,1996.
[35]李连发,王劲峰,刘纪远.国土遥感调查的空间抽样优化决策[J].中国科学(D),2004,(10):975-982.
[36] Wang J F, Zhuang D F and Li L F. Spatial sampling design for monitoring the area ofcultivated land[J]. International Journal of Remote Sensing,2002,13(2):263-284.
[37] Webster R Welhama S J and Pottsb J M, et al. Estimating the spatial scales of regionalizedvariables by nested sampling, hierarchical analysis of variance and residual maximumlikelihood[J]. Computers&Geosciences,2006,32(9):1320-1333.
[38] Simbahan G C and Dobermaan A. Sampling optimization based on secondary informationand its utilization in soil carbon mapping[J]. Geoderma,2006,133(3-4):345-362.
[39] Brus D J and Gruijter J J. Random sampling or geostatistical modelling? Choosing betweendesign-based and model-based sampling strategies for soil (with discussion)[J]. Geoderma,1997,80(1-2):1-44.
[40] Rogerson P A, Delmelle E and Batta R. Optimal sampling design for variables with varyingspatial importance[J]. Geographical Analysis,2004,36(2):177-194.
[41] Journel A G. Nonparametric geostatistics for risk and additional sampling assessment. In:Principles of Environmental Sampling[J]. American Chemical Society,1988:45-72.
[42] Van Groenigen J W, Siderius W and Stein A. Constrained optimisation of soil sampling forminimisation of the kriging variance[J]. Geoderma,1999,87(3-4):239-259.
[43] Van Groenigen J W, Pieters G and Stein A. Optimizing spatial sampling for multivariatecontamination in urban areas[J]. Environmetrics,2000,11(2):227-244.
[44] Stevens Jr D L. Variable density grid-based sampling designs for continuous spatialpopulations[J]. Environmetrics,1997,8(3):167-195.
[45] Smith D R, Conroy M J and Brakhage D H. Efficiency of adaptive cluster sampling forestimating density of wintering waterfowl[J]. Biometrics,1995,51(2):777-788
[46] Bertolino F, Luciano A and Racugno W. Some aspects of detection networks optimizationwith the kriging procedure[J]. Metron,1983,41(3):91-107.
[47] Van Groenigen J W, Siderius W and Stein A. Constrained optimisation of soil sampling forminimisation of the kriging variance[J]. Geoderma,1999,87(3-4):239-259.
[48] Nunes L M. Optimal estuarine sediment monitoring network design with simulatedannealing[J]. Journal of Environmental Management,2006,78(3):294-304.
[49] Zio S D, Fontanella L and Ippoliti L. Optimal spatial sampling schemes for environmentalsurveys[J], Environmental and Ecological Statistics,2004,11(4):397-414.
[50] Stevens Jr. D L. Spatial properties of design-based versus model-based approaches toenvironmental sampling. In: Caetano M, Painho M(eds.).7th International Symposium onSpatial Accuracy Assessment in Natural Resources and Environmental Sciences,2006.
[51] Lu D and Weng Q. A survey of image classification methods and techniques for improvingclassification performance[J]. International Journal of Remote Sensing,2007,28(5):823-870.
[52] Gong P and Howarth J P. Frequency-based contextual classification and gray-level vectorreduction for land-use identification[J]. Photogrammetric Engineering and Remote Sensing,1992,8(4):423-437.
[53] Foody G M. Approaches for the Production and Evaluation of Fuzzy Land CoverClassifications from Remotely sensed Data[J]. International Journal of Remote Sensing,1996,17(7):1317-1340.
[54]李祚泳.用BP神经网络实现多波段遥感图像的监督分类[J].红外与毫米波学报,1998,17(2):153-156.
[55] Gallego F J. Remote sensing and land cover area estimation[J]. International Journal ofRemote Sensing,2004,25(15):3019-3047.
[56]术洪磊,毛赞猷. GIS辅助下的基于知识的遥感影像分类方法研究—以土地覆盖/土地利用类型为例[J].测绘学报,1997,26(4):328-336.
[57]余凡,李海涛,万紫.结合贝叶斯理论和MRF的主被动遥感数据协同分类[J].遥感学报,2012,16(4):809-825.
[58] Zhong, Y and Zhang, L. An adaptive artificial immune network for supervised classificationof multi/hyper-spectral remote sensing imagery[J]. IEEE Transactions on Geoscience andRemote Sensing,2012,50(3):894-909.
[59] Kavzoglu T and Mather P M. The use of back propagating artificial neural networks in landcover classification[J].International Journal of Remote Sensing,2003,24(23):4907-4938.
[60] Huang C, Davis L S and Townshend J R G. An assessment of support vector machines forland cover classification[J]. International Journal of Remote Sensing,2002,23(4):725-749.
[61] Zhu G and Blumberg D G. Classification using ASTER data and SVM algorithms: the casestudy of Beer Sheva Israel[J], Remote Sensing of Environment,2002,80(2):233-240.
[62] Pal M. Support vector machine-based feature selection for land cover classification: a casestudy with DAIS hyperspectral data[J], International Journal of Remote Sensing,2006,27(14):2877-2894.
[63] Pal M and Mather P M. Assessment of the effectiveness of support vector machines forhyperspectral data[J]. Future Generation Computer Systems,2004,20(7):1215-1225.
[64] Melgani F and Bruzzone L. Classification of hyperspectral remote sensing images withSupport Vector Machines[J], IEEE Transactions on Geoscience and Remote Sensing,2004,42(8):1778-1790.
[65] Hodgson M E, Jensen J R and Tullis J A. et al. Synergistic use lidar and color aerialphotography for mapping urban parcel imperviousness[J]. Photogrammetric Engineering andRemote Sensing,2003,69(9):973-980.
[66]杜培军.遥感原理与应用[M].徐州:中国矿业大学出版社,2006:169.
[67] Foody G M and Mathur A. A relative evaluation of multiclass image classification by supportvector machines[J]. IEEE Transactions on Geoscience and Remote Sensing,2004,42(6):1335-1334.
[68] Paola J D and Schowenger R A. A detailed comparison of back propagation neural networkand maximum likelihood classifiers for urban land use classification[J]. IEEE Transactionson Geoscience and Remote Sensing,1995,33(4):981-996.
[69]李强,王正志.基于人工神经网络和经验知识的遥感信息分类综合方法[J].自动化学报,2000,26(2):233-238.
[70]骆剑承,周成虎,杨艳.基于径向基函数(RBF)映射理论的遥感影像分类模型研究[J].中国图象图形学报,2000,5(2):94-994.
[71]王耀南.小波神经网络的遥感图像分类[J].中国图象图形学报,1999,4(5):368-371.
[72] Benediktsson J A, Swain P H and Ersoy O K. Neural network approaches versus statisticalmethods in classification of multisource remote sensing data[J]. IEEE Transactions onGeoscience and Remote Sensing,1990,28(4):540-552.
[73] Murthy C S, Raju P V and Badrinath K V S. Classification of wheat crop with multi-temporalimages: performance of maximum likelihood and artificial neural networks[J]. InternationalJournal of Remote Sensing,2003,24(23):4871-4890.
[74] Heerman P D and Khazenie N. Classification of multispectral remote sensing data using aback-propagation neural network[J]. IEEE Transactions on Geoscience and Remote Sensing,1992,30(1):81-88.
[75]张利,吴华玉,卢秀颖.基于粗糙集的改进BP神经网络算法研究[J].大连理工大学学报,2009,49(6):971-976.
[76] Vapnik V. N. The Nature of Statistical Learning Theory. New York: Springer,1995:114.
[77] Camps-Valls G, Gómez-Chova L and Calpe-Maravilla J, et al. Robust support vector methodfor hyperspectral data classification and knowledge discovery[J]. IEEE Transactions onGeoscience and Remote Sensing,2004,42(7),1530-1542.
[78] Melgani F and Bruzzone L. Classification of hyperspectral remote sensing images withSupport Vector Machines[J], IEEE Transactions on Geoscience and Remote Sensing,2004,42(8):1778-1790.
[79] Dixon B and Candade N. Multispectral land use classification using neural networks andsupport vector machines: one or the other, or both?[J], International Journal of RemoteSensing,2008,29(4):1185-1206.
[80] Roli F and Fumera G. Support vector machines for remote-sensing image classification[C], inProc. EOS/SPIE Symposium,(Barcelona),2000.
[81] Hsu C W and Lin C J. A comparison of methods for multiclass support vector machines[J].IEEE Transactions on Neural Networks,2002,13(2):415-425.
[82] Platt J C, Cristianini N and Shawe-Taylor J. Large margin DAG's for multi-classclassification[C]. Advances in Neural Information Processing Systems,2000,12:547-553.
[83] Bruzzone L, Mingmin C and Marconcini M. A novel transductive SVM for semisupervisedclassification of remote-sensing images[J]. IEEE Transactions on Geoscience and RemoteSensing,2006,44(11):3363-3373.
[84] Demir B and Erturk S. Hyperspectral image classification using relevance vector machines[J].IEEE Transactions on Geoscience and Remote Sensing,2007,4(4):586-590.
[85] Foody G M. RVM-based multi-class classification of remotely sensed data[J]. InternationalJournal of Remote Sensing,2008,29(6):1817-1823.
[86] Zhang R and Ma J. An improved SVM method P-SVM for classification of remotely senseddata[J]. International Journal of Remote Sensing,2008,29(20):6029-6036.
[87] Quinlan J. Induction of decision trees[J]. Machine learning,1986,1(1):81-106.
[88] Friedl M and Brodley C. Decision tree classification of land cover from remotely senseddata[J]. Remote Sensing of Environment,1997,61(3):399-409.
[89] Friedl M, Brodley C and Strahler A. Maximizing land cover classification accuraciesproduced by decision trees at continental to global scales[J]. IEEE Transactions onGeoscience and Remote Sensing,1999,37(2):969-977.
[90] Hansen M, Dubayah R and Defries R. Classification trees: an alternative to traditional landcover classifiers[J], International Journal of Remote Sensing,1996,17(5):1075-1081.
[91] Belward A S and De Hoyos A. A comparison of supervised maximum likelihood and decisiontree classification for crop cover estimation from multitemporal LANDSAT MSS data[J],International Journal of Remote Sensing,1987,8(2):229-235.
[92] Wardlow B D and Egbert S L. Large-area crop mapping using time-series MODIS250mNDVI data: An assessment for the US Central Great Plains[J], Remote Sensing ofEnvironment,2008,112(3):1096-1116.
[93] Schneider A, McIver D K and Friedl M A, et al. Classification of urban areas at continentalscales using remotely sensed data[C], IEEE Geoscience and Remote Sensing Symposium,2001,5:2146-2148.
[94]吴成东,许可,韩中华,裴涛.基于粗糙集和决策树的数据挖掘方法[J].东北大学学报,2006,27(5):481-484.
[95]乔梅,韩文秀.基于Rough集的决策树算法[J].天津大学学报,2005,38(9):842-846.
[96] Asit Kumar Das and Jaya Sil. An efficient classifier design integrating rough set and setoriented database operations[J]. Applied Soft Computing,2011,11(2):2279-2285.
[97] Stefanov W L, Ramsey M S and Christensen P R. Monitoring urban land cover change: Anexpert system approach to land cover classification of semiarid to arid urban centers[J],Remote Sensing of Environment,2001,77(2):173-185
[98]贾海峰,刘雪华.环境遥感原理与应用[M],北京:清华大学出版社,2006:204.
[99]甘淑,袁希平,何大明.遥感专家分类系统在滇西北植被信息提取中的应用试验研究[J].云南大学学报(自然科学版),2003,25(6):553-557.
[100] Cohen, Y and Shoshany M. Integration of remote sensing, GIS and expert knowledge innational knowledge-based crop recognition in Mediterranean environment[J], InternationalJournal of Applied Earth Observation and Geoinformation,2002,4(1):75-87.
[101] Schmidt K S, Skidmore A K and Kloosterman E H, et al. Mapping coastal vegetation usingan expert system and hyperspectral imagery[J]. Photogrammetric Engineering and RemoteSensing,2004,70(6):703-715.
[102] Lucas R, Rowlands A and Brown A, et al. Rule-based classification of multi-temporalsatellite imagery for habitat and agricultural land cover mapping[J]. ISPRS Journal ofPhotogrammetry&Remote Sensing,2007,62(3):165-185.
[103] Zadeh L A. Fuzzy Sets. Information and Control[J].1965,8(3):338-353
[104] Liu Y M and Luo M K. Fuzzy Topology[M]. World Scientific,1997.
[105] Wu G and Zheng C. Fuzzy boundary and characteristic properties oforder-homomorphisms[J]. Fuzzy Sets and Systems1991,39(3):329-337.
[106] Tang X M, Kainz W and Fang Y. Reasoning about changes of land covers with fuzzysettings[J]. International Journal of Remote Sensing,2005,26(14):3025-3046.
[107] Andrew J T, Hugh G L and Peter M A, et al. Super-resolution target identification fromremotely sensed images using a hopfield neural network[J]. IEEE Transactions OnGeoscience and Remote Sensing,2001,39(4):781-796.
[108] Kittler J. Improving recognition rates by classifier combination: a theoretical Framework
[C]. Progress in Handwriting Recognition. Singapore: World Scientific Publishing,1997:231-248.
[109] Xu L, Krzyzak A and Suen C Y. Methods of combining multiple classifiers and theirapplications to handwriting recognition[J]. IEEE Transactions on Systems, Man andCybernetics,1992,22(3):418-435.
[110] Ho T K, Hull J J and Srihari S N. Decision combination in multiple classifier systems[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1994,16(1):66-75.
[111] Cho S B and Kim J H. Multiple network fusion using fuzzy logic[J]. IEEE Transactions onNeural Networks,1995,6(2):497-501.
[112] Lam L and Suen C Y. Optimal combination of pattern Classifiers[J]. Pattern RecognitionLetters,1995,6(9):945-954.
[113] Freund Y and Schapire R E. Experiments with a new Boosting algorithm. In: LorenzaSAITTA, ed. Machine Learning: Proceedings of the Thirteenth International Conference(ICML '96).1996, San Francisco, CA: Morgan Kaufmann:148-156.
[114] Breiman L. Bagging Predictors[J]. Machine Learning,1996,24(2):123-140.
[115] Lam L and Suen C Y. Application of majority voting to pattern recognition: An analysis ofits behaviour and performance[J]. IEEE Transactions on System, Man, and Cybernetics,Part A,1997,27(5):553-568.
[116] Kittler J, Hatef M and Duin R P, et al. On combining classifiers [J]. IEEE Transactions onPattern Analysis and Machine Intelligence,1998,20(3):226-239.
[117] Opitz D and Maclin R. Popular ensemble methods: an empirical study[J]. Journal ofArtificial Intelligence Research,1999,11(1):169-198.
[118] Jain A K, Duin R P W and Mao J-C. Statistical pattern recognition: A review[J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2000,22(1):4-37.
[119] Skurichina M and Duin R P W. Bagging, Boosting and the random dubspace method forlinear classifiers [J]. Pattern Analysis and Applications,2002,5(2):121-135.
[120] Zhou Z H and Tang W. Selective ensemble of decision trees[J]. Lecture Notes in ComputerScience,2003,2639:476-483.
[121] Melville P and Mooney R J. Creating diversity in ensembles using artificial data[J].Information Fusion,2005,6(1):99-111.
[122]肖旭红,戴汝为.一种识别手写汉字的多分类器集成方法[J].自动化学报,1997,23(5):621-627.
[123]王岚,陈坷,迟惠生.多特征组合多分类器的方法用于“文本无关”的说话人辨认[J].北京大学学报(自然科学版),1998,34(2-3):143-I50.
[124] Zhou Z H and Yu Y. Adapt bagging to nearest neighbor classifiers[J]. Journal of ComputerScience and Technology,2005(1):48-54.
[125] Steele B M. Combining multiple classifiers: an application using spatial and remotelysensed information for land cover type mapping[J]. Remote Sensing of Environment,2000,74(3):545-556.
[126] Huang Z and Lees B G. Combining Non-Parametric Models for Multisource PredictiveForest Mapping[J]. Photogrammetric Engineering and Remote Sensing,2004,70(4):415-425.
[127] Foody G M, Boyd D S and Sanchez-Hernandez C. Mapping a specific class with anensemble of classifiers[J]. International Journal of Remote Sensing,2007,28(7):1733-1746.
[128] Serra P and Pons X. Monitoring farmers' decisions on Mediterranean irrigated crops usingsatellite image time series[J]. International Journal of Remote Sensing,2008,29(8):2293-2316.
[129] Maulik U and Chakraborty D A. A robust multiple classifier system for pixel classificationof remote sensing images[J]. Fundamenta Informaticae,2010,101(4):286-304.
[130] Zhang Y. Optimisation of building detection in satellite images by combining multispectralclassification and texture filtering[J]. ISPRS Journal of Photogrammetry and RemoteSensing,1999,54(1):50-60.
[131] Zhang Y. Texture-integrated classification of urban treed areas in High-ResolutionColor-Infrared imagery[J]. Photogrammetric Engineering and Remote Sensing,2001,67(12):1359-1365
[132]吴昊.综合纹理特征的高光谱遥感图像分类方法[J].计算机工程与设计,2012,33(5):1993-2006
[133]李厚强,刘政凯,林峰.基于分形理论的航空图像分类方法[J].遥感学报,2001,5(5):353-357.
[134] Benediktsson J A, Palmason J A and Sveinsson J R. Classification of hyperspectral datafrom urban areas based on extended morphological profiles[J]. IEEE Transactions onGeoscience and Remote Sensing,2005,43(3):480-491.
[135] Fauvel, M, Benediktsson J A and Chanussot J, et al. Spectral and spatial classification ofhyperspectral data using SVMs and morphological profiles[J]. IEEE Transactions onGeoscience and Remote Sensing,2005,43(3):3804-3815.
[136]汪闽,骆剑承,周成虎等.结合高斯马尔可夫随机场纹理模型与支撑向量机在高分辨率遥感图像上提取道路网[J].遥感学报,2005,9(3):271-276.
[137] Zhao Y D, Zhang L P and Li P X, et al. classification of high spatial resolution imageryusing improved gaussian markov random-field-based texture features[J]. IEEE Transactionson Geoscience and Remote Sensing,2007,45(5):1458-1468.
[138] Segl K, Reossner S and Heiden U, et al. Fusion of spectral and shape features foridentification of urban surface cover types using reflective and thermal hyperspectral data[J].ISPRS Journal of Photogrammetry and Remote Sensing,2003,58(1-2):99-112.
[139] Shackelford A and Davis X H. Hierarchical fuzzy classification approach for high-resolutionmultispectral data over urban areas[J]. IEEE Transactions on Geoscience and RemoteSensing,2003,41(9):1920-1932.
[140] Zhang L P, Huang X and Huang B, et al. A pixel shape index coupled with spectralinformation for classification of high spatial resolution remotely sensed imagery[J]. IEEETransactions on Geoscience and Remote Sensing,2006,44(10):2950-2961.
[141]黄昕,张良培,李平湘.融合形状和光谱的高空间分辨率遥感影像分类[J].遥感学报,2007,11(2):193-200.
[142] Huang X Zhang L P and Li P X, et al. classification and extraction of spatial features inurban areas using high-resolution multispectral imagery[J]. IEEE Geoscience and RemoteSensing Letters,2007,4(2):260-264.
[143] Huang X Zhang L P and Li P X. Classification of very high spatial resolution imagery basedon the fusion of edge and multispectral information[J]. Photogrammetric Engineering andRemote Sensing,2008,74(12):1585-1596.
[144] Congalton R G and Mead R A. A quantitative method to test for consistency and correctnessin photointerpretation[J]. Photogrammetric Engineering and Remote Sensing,1983,49(1):69-74.
[145] Aronoff S. The minimum accuracy value as an index of classification accuracy[J].Photogrammetric Engineering and Remote Sensing,1985,51(1):99-111.
[146] Koukoulas S and Blackburn G A. Introducing new indices for accuracy evaluation ofclassified images represnting semi-natural woodland environments[J]. PhotogrammetricEngineering and Remote Sensing,2001,67(4):499-510.
[147] Woodcock C E and Gopal S. Fuzzy set theory and thematic maps: accuracy assessment andarea estimation[J]. International Journal of Geographical Information Science,2000,14(2):153-172.
[148] Congalton R G, Oderwald R G and Mead R A. Assessing landsat classification accuracyusing discrete multivariate statistical techniques[J]. Photogrammetric Engineering andRemote Sensing,1983,49(12):1671-1678.
[149] Canters F. Evaluating the uncertainty of area estimates derived from fuzzy land-coverclassification[J]. Photogrammetric Engineering and Remote Sensing,1997,63(4):403–414.
[150] Foody G M. Status of land cover classification accuracy assessment[J]. Remote Sensing ofEnvironment,2002,80(1):185-201.
[151]柏延臣,王劲峰.遥感数据专题分类不确定性评价研究:进展、问题与展望[J].地球科学进展,2005,20(11):1218-1225.
[152] Foody G M, Campbell N A and Wood T F. Derivation and applications of probabilisticmeasures of class membership from the maximum likelihood classification[J].Photogrammetric Engineering and Remote Sensing,1992,58(9):1335-1341.
[153] Stephen V S and Raymond L C. Design and analysis for thematic map accuracy assessment:fundamental principles[J]. Remote Sensing of Environment,1998,64(3):331-344.
[154] Stephen V S. Sampling designs for assessing map accuracy[C]. Proceedings of the8thInternational Symposium on Spatial Accuracy Assessment in Natural Resources andEnvironmental Sciences, Shanghai, P. R. China,2008:8-15.
[155] Ge Y, li S P and Lakhan V C, et al. Exploring uncertainty in remotely sensed data withparallel coordinate plots[J]. International Journal of Applied Earth Observation andGeoinformation,2009,11(6):413-422.
[156] Rosenfield G H. Analysis of variance of thematic mapping experiment data[J],Photogrammetric Engineering and Remote Sensing,1981,47(12):1685-1692.
[157] Goodchild M F, Sun G Q and Yang S R. Development and test of and error model forcategorical data[J]. International Journal of Geographic Information Systems,1992,6(2):87-104.
[158] Congalton R G. A review of assessing the accuracy of classifications of remotely senseddata[J]. Remote Sensing of Environment,1991,37(1):35-46.
[159] Stevens Don L J R. Variable density grid-based sampling designs for continuous spatialpopulations[J]. Environmetrics,1997,8(3):164-195.
[160] Foody G M and Mathur A. The use of small training sets containing mixed pixels foraccurate hard image classification: Training on mixed spectral responses for classificationby a SVM [J]. Remote Sensing of Environment,2006,103(2):179-189.
[161] Li Z L and Huang P Z. Quantitative measures for spatial information of maps[J].International Journal of Geographical Information Science,2002,16(7):699-709.
[162] Cortes C and Vapnik V N. Support vector networks[J]. Machine Learning,1995,20:273-297.
[163] Boser B E, Guyon I M and Vapnik V N. A training algorithm for optimal marginclassifiers[C]. In Proc. Fifth Ann. Workshop on Comp. Learn Theory., Pittsburgh, PA,1992:144-152.
[164] Hsu C W and Lin C J. A comparison of methods for multi-class support vector machines[J].IEEE Transactions on Neural Networks,2002,13(2):415-425.
[165] Platt J C. Probabilities for Support Vector Machines[M]. Massachusetts: MIT Press,2000:61-74.
[166] Wu T F, Lin CH J and Weng R C. Probability estimates for multi-class classification bypairwise coupling[J]. Journal of Machine Learning Research,2004,5:975-1005.
[167] Chang C L. Fuzzy topological spaces[J]. Journal of Mathematical Analysis and Applications,1968,24(2):182-190.
[168] Liu K F and Shi W Z. Computation of fuzzy topological relations of spatial objects based oninduced fuzzy topology[J], International Journal of Geographical Information Systems,2006,20(8):857-883.
[169] Shi W Z and Liu K F. A fuzzy topology for computing the interior, boundary, and exterior ofspatial objects quantitatively in GIS[J]. Computers&Geosciences,2007,33(7):898-915.
[170] Shi W Z, Liu K F and Huang C Q. A fuzzy Topology-integrated area object extractionmethod[J]. IEEE Transactions on Geoscience and Remote Sensing,2010,48(1):147-154.
[171] Hsu C W, Chang C C and Lin C J. A practical guide to support vector classification,2010.
[Online]. Available:
[172] LekasK B, Likask A and Ga latsanos N P. A spatially constrained mixture model for imagesegmentation [J]. IEEE Transactions on Neural Networks,2005,16(2):494-498.
[173] Bezdek J C. Pattern recognition with fuzzy objective function algorithms [M]. New York:Plenum Press,1981:3-5.
[174]王顺凤，张建伟.基于Gibbs场与模糊C-均值聚类的脑MR图像分割[J].计算机应用,2008,28(7):1750-1752.
[175] LI S Z. Markov random field modeling in image analysis (Third Edition)[M]. Berlin:Springer-Verlag,2009:65-67.
[176] GC, C., J. AK. Markov random field texture models[J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,1983,1(1):25-39.
[177]杨龙平,丁宣浩,孙建明.一种新的基于小波变换的图像消噪方法[J].云南民族大学学报:自然科学版,2005,14(1):71-74.
[178] Suzuki T, Sugiyama M and Kanamori T, et al. Mutual information estimation reveals globalassociations between stimuli and biological processes[J]. BMC Bioinformatics,2009,10(1):1391-1445.
[179] Chang C C and Lin C J. LIBSVM: A Library for Support Vector Machines,2001.[Online].Available: http://www.csie.ntu. edu.tw/~cjlin/libsvm/.
[180] Theobald, D.M. and J.B. Norman.2006. Spatially-balanced sampling using The ReversedRandomized Quadrant-Recursive Raster algorithm: A User’s Guide for the RRQRR ArcGISv9.1tool. URL: http://www.nrel.colostate.edu/projects/starmap.
[181] Biernacki C. Initializing EM using the properties of its trajectories in Gaussian Mixtures[J].Statistics and Computing,2004,14(3):267-279.
[182] Fraley C. Algorithms for model-based Gaussian hierarchical clustering[J]. SIAM Journal onScientific Computing archive,1998,20(1):270–281.
[183] Fraley C and Raftery A E. Model-based clustering, discriminant analysis and densityestimation[J]. Journal of the American Statistical Association,2002,97(458):611–631.
[2]史文中,陈江平,詹庆明等.可靠性空间分析初探[J].武汉大学学报(信息科学版),2012,37(8):883-991.
[3] Congalton R G. A review of assessing the accuracy of classifications of remotely senseddata[J]. Remote Sensing of Environment,1991,37(1):35-46.
[4] Shi W Z. Principle of modeling uncertainties in spatial data and spatial analyses[M]. New York:Taylor&Francis Group/CRC Press,2009:85.
[5](美)延森(Jensen, J R)著;陈晓玲等译.遥感数字影像处理导论(原书第3版)[M].北京:机械工业出版社,2007:325-326.
[6] Han J and Kamber M. Data mining: concepts and techniques[M]. Los Altos, CA, USA:Morgan Kaufmann Publishers,2001:142.
[7] Hixson M, Scholz D, Fuhs N and Akiyama T. Evaluation of several schemes for classificationof remotely sensed data[J]. Photogrammetric Engineering and Remote Sensing,1990,46(1)：1547-1553.
[8] Jackson Q and Landgrebe D A. An adaptive classifier design for high-dimensional dataanalysis with a limited training data set[J]. IEEE Transactions on Geoscience and RemoteSensing,2001,39(12):2664-2679.
[9] Nelson R F, Latty R S and Mott G. Classifying northern forests using Thematic Mappersimulator data[J]. Photogrammetric Engineering and Remote Sensing,1984,50(5):607-617.
[10] Van Niel T G, McVicar T R and Datt B. On the relationship between training sample size anddata dimensionality of broadband multi-temporal classification[J]. Remote Sensing ofEnvironment,2005,98(4):468-480.
[11] Zhuang X, Engel B A and Lozano-Garcia D F, et al. Optimisation of training data required forneuro-classification[J]. International Journal of Remote Sensing,1994,15(16):3271-3277.
[12] Foody G M and Arora M K. An evaluation of some factors affecting the accuracy ofclassification by an artificial neural network[J]. International Journal of Remote Sensing,1997,18(4):799-810.
[13]阎静,王汶,李湘阁.利用神经网络方法提取水稻种植面积-以湖北省双季早稻为例[J].遥感学报,2001,5(3):227-231.
[14] Foody G M, Mathur A. Toward intelligent training of supervised image classifications:directing training data acquisition for SVM classification[J]. Remote Sensing of Environment,2004,93(1-2):107-117.
[15] Foody G M. Thematic map comparison: Evaluating the statistical significance of differencesin classification accuracy[J]. Photogrammetric Engineering and Remote Sensing,2004,70(5):627-633.
[16] Foody G M, Mathur A, Sanchez-Hernandez C and Boyd D S. Training set size requirementsfor the classification of a specific class[J]. Remote Sensing of Environment,2006,104(1):1-14.
[17]朱秀芳,潘耀忠,张锦水等.训练样本对TM尺度小麦种植面积测量精度影响研究(Ⅰ)-训练样本与分类方法间分类精度响应关系研究[J].遥感学报,2007,11(6):826-837.
[18]薄树奎,丁琳.训练样本数目选择对面向对象影像分类方法精度的影响[J].中国图象图形学报,2010,15(7):1106-1111.
[19] Kuo B C and Landgrebe D A. A covariance estimator for small sample size classificationproblems and its application to feature extraction[J]. IEEE Transactions on Geoscience andRemote Sensing,2002,40(4):814-819.
[20] Arai K. A supervised thematic mapper classification with a purification of training samples.International Journal of Remote Sensing,1992,13(11):2039-2049.
[21]吴健平,杨星卫.遥感数据监督分类中训练样本的纯化[J].国土资源遥感,1996,27(1):36-41.
[22] Foody G M and Mathur A. The use of small training sets containing mixed pixels foraccurate hard image classification: Training on mixed spectral responses for classification bya SVM [J]. Remote Sensing of Environment,2006,103,(2),179-189.
[23] Schmidt K S and Skidmore A K. Spectral discrimination of vegetation types in a coastalwetland[J]. Remote Sensing of Environment,2003,85(1):92-108.
[24]杜红艳,张洪岩,张正祥. GIS支持下的湿地遥感信息高精度分类方法研究[J].遥感技术与应用,2004,19(4):244-248.
[25]连石柱.基于分离度的图像特征提取于识别[J].中国图象图形学报,1996,1(3):196-200.
[26]张治英,徐德忠,周云等.不同特征遥感图像在江滩钉螺孳生地监测中的应用[J].西安交通大学学报,2004,25(3):304-307.
[27] Key T, Warner T A and McGraw J B, et al. A comparison of multispectral and multitemporalinformation in high spatial resolution imagery for classification of individual tree species in atemperate hardwood forest[J]. Remote Sensing of Environment,2001,75(1):100-112.
[28] Koukoulas S and George Alan B. Introducing new indices for accuracy evaluation ofclassified images representing semi-natural woodland environments[J]. PhotogrammetricEngineering and Remote Sensing,2001,67(4):499-510.
[29] Van Niela T G, McVicarb T R and Dattc B. On the relationship between training sample sizeand data dimensionality: Monte Carlo analysis of broadband multi-temporal classification[J].Remote Sensing of Environment,2005,98(4):468-480.
[30] Scott. C T. Sampling method for estimating change in forest resources[J]. EcologicalApplications,1998,8(2):228-233.
[31] Dunn R and Harrison A R. Two-dimensional systematic sampling of land use[J]. AppliedStatistics,1993,42(4):585-601.
[32]曹志冬,王劲峰,李连发等.地理空间中不同分层抽样方式的分层效率与优化策略[J].地理科学进展,2008,27(3):152-160.
[33]冯士雍.抽样调查应用与理论中的若干前沿问题[J].统计与信息论坛,2007,22(1):5-31.
[34]冯士雍,施锡铨.抽样调查:理论、方法与实践[M].上海:上海科学技术出版社,1996.
[35]李连发,王劲峰,刘纪远.国土遥感调查的空间抽样优化决策[J].中国科学(D),2004,(10):975-982.
[36] Wang J F, Zhuang D F and Li L F. Spatial sampling design for monitoring the area ofcultivated land[J]. International Journal of Remote Sensing,2002,13(2):263-284.
[37] Webster R Welhama S J and Pottsb J M, et al. Estimating the spatial scales of regionalizedvariables by nested sampling, hierarchical analysis of variance and residual maximumlikelihood[J]. Computers&Geosciences,2006,32(9):1320-1333.
[38] Simbahan G C and Dobermaan A. Sampling optimization based on secondary informationand its utilization in soil carbon mapping[J]. Geoderma,2006,133(3-4):345-362.
[39] Brus D J and Gruijter J J. Random sampling or geostatistical modelling? Choosing betweendesign-based and model-based sampling strategies for soil (with discussion)[J]. Geoderma,1997,80(1-2):1-44.
[40] Rogerson P A, Delmelle E and Batta R. Optimal sampling design for variables with varyingspatial importance[J]. Geographical Analysis,2004,36(2):177-194.
[41] Journel A G. Nonparametric geostatistics for risk and additional sampling assessment. In:Principles of Environmental Sampling[J]. American Chemical Society,1988:45-72.
[42] Van Groenigen J W, Siderius W and Stein A. Constrained optimisation of soil sampling forminimisation of the kriging variance[J]. Geoderma,1999,87(3-4):239-259.
[43] Van Groenigen J W, Pieters G and Stein A. Optimizing spatial sampling for multivariatecontamination in urban areas[J]. Environmetrics,2000,11(2):227-244.
[44] Stevens Jr D L. Variable density grid-based sampling designs for continuous spatialpopulations[J]. Environmetrics,1997,8(3):167-195.
[45] Smith D R, Conroy M J and Brakhage D H. Efficiency of adaptive cluster sampling forestimating density of wintering waterfowl[J]. Biometrics,1995,51(2):777-788
[46] Bertolino F, Luciano A and Racugno W. Some aspects of detection networks optimizationwith the kriging procedure[J]. Metron,1983,41(3):91-107.
[47] Van Groenigen J W, Siderius W and Stein A. Constrained optimisation of soil sampling forminimisation of the kriging variance[J]. Geoderma,1999,87(3-4):239-259.
[48] Nunes L M. Optimal estuarine sediment monitoring network design with simulatedannealing[J]. Journal of Environmental Management,2006,78(3):294-304.
[49] Zio S D, Fontanella L and Ippoliti L. Optimal spatial sampling schemes for environmentalsurveys[J], Environmental and Ecological Statistics,2004,11(4):397-414.
[50] Stevens Jr. D L. Spatial properties of design-based versus model-based approaches toenvironmental sampling. In: Caetano M, Painho M(eds.).7th International Symposium onSpatial Accuracy Assessment in Natural Resources and Environmental Sciences,2006.
[51] Lu D and Weng Q. A survey of image classification methods and techniques for improvingclassification performance[J]. International Journal of Remote Sensing,2007,28(5):823-870.
[52] Gong P and Howarth J P. Frequency-based contextual classification and gray-level vectorreduction for land-use identification[J]. Photogrammetric Engineering and Remote Sensing,1992,8(4):423-437.
[53] Foody G M. Approaches for the Production and Evaluation of Fuzzy Land CoverClassifications from Remotely sensed Data[J]. International Journal of Remote Sensing,1996,17(7):1317-1340.
[54]李祚泳.用BP神经网络实现多波段遥感图像的监督分类[J].红外与毫米波学报,1998,17(2):153-156.
[55] Gallego F J. Remote sensing and land cover area estimation[J]. International Journal ofRemote Sensing,2004,25(15):3019-3047.
[56]术洪磊,毛赞猷. GIS辅助下的基于知识的遥感影像分类方法研究—以土地覆盖/土地利用类型为例[J].测绘学报,1997,26(4):328-336.
[57]余凡,李海涛,万紫.结合贝叶斯理论和MRF的主被动遥感数据协同分类[J].遥感学报,2012,16(4):809-825.
[58] Zhong, Y and Zhang, L. An adaptive artificial immune network for supervised classificationof multi/hyper-spectral remote sensing imagery[J]. IEEE Transactions on Geoscience andRemote Sensing,2012,50(3):894-909.
[59] Kavzoglu T and Mather P M. The use of back propagating artificial neural networks in landcover classification[J].International Journal of Remote Sensing,2003,24(23):4907-4938.
[60] Huang C, Davis L S and Townshend J R G. An assessment of support vector machines forland cover classification[J]. International Journal of Remote Sensing,2002,23(4):725-749.
[61] Zhu G and Blumberg D G. Classification using ASTER data and SVM algorithms: the casestudy of Beer Sheva Israel[J], Remote Sensing of Environment,2002,80(2):233-240.
[62] Pal M. Support vector machine-based feature selection for land cover classification: a casestudy with DAIS hyperspectral data[J], International Journal of Remote Sensing,2006,27(14):2877-2894.
[63] Pal M and Mather P M. Assessment of the effectiveness of support vector machines forhyperspectral data[J]. Future Generation Computer Systems,2004,20(7):1215-1225.
[64] Melgani F and Bruzzone L. Classification of hyperspectral remote sensing images withSupport Vector Machines[J], IEEE Transactions on Geoscience and Remote Sensing,2004,42(8):1778-1790.
[65] Hodgson M E, Jensen J R and Tullis J A. et al. Synergistic use lidar and color aerialphotography for mapping urban parcel imperviousness[J]. Photogrammetric Engineering andRemote Sensing,2003,69(9):973-980.
[66]杜培军.遥感原理与应用[M].徐州:中国矿业大学出版社,2006:169.
[67] Foody G M and Mathur A. A relative evaluation of multiclass image classification by supportvector machines[J]. IEEE Transactions on Geoscience and Remote Sensing,2004,42(6):1335-1334.
[68] Paola J D and Schowenger R A. A detailed comparison of back propagation neural networkand maximum likelihood classifiers for urban land use classification[J]. IEEE Transactionson Geoscience and Remote Sensing,1995,33(4):981-996.
[69]李强,王正志.基于人工神经网络和经验知识的遥感信息分类综合方法[J].自动化学报,2000,26(2):233-238.
[70]骆剑承,周成虎,杨艳.基于径向基函数(RBF)映射理论的遥感影像分类模型研究[J].中国图象图形学报,2000,5(2):94-994.
[71]王耀南.小波神经网络的遥感图像分类[J].中国图象图形学报,1999,4(5):368-371.
[72] Benediktsson J A, Swain P H and Ersoy O K. Neural network approaches versus statisticalmethods in classification of multisource remote sensing data[J]. IEEE Transactions onGeoscience and Remote Sensing,1990,28(4):540-552.
[73] Murthy C S, Raju P V and Badrinath K V S. Classification of wheat crop with multi-temporalimages: performance of maximum likelihood and artificial neural networks[J]. InternationalJournal of Remote Sensing,2003,24(23):4871-4890.
[74] Heerman P D and Khazenie N. Classification of multispectral remote sensing data using aback-propagation neural network[J]. IEEE Transactions on Geoscience and Remote Sensing,1992,30(1):81-88.
[75]张利,吴华玉,卢秀颖.基于粗糙集的改进BP神经网络算法研究[J].大连理工大学学报,2009,49(6):971-976.
[76] Vapnik V. N. The Nature of Statistical Learning Theory. New York: Springer,1995:114.
[77] Camps-Valls G, Gómez-Chova L and Calpe-Maravilla J, et al. Robust support vector methodfor hyperspectral data classification and knowledge discovery[J]. IEEE Transactions onGeoscience and Remote Sensing,2004,42(7),1530-1542.
[78] Melgani F and Bruzzone L. Classification of hyperspectral remote sensing images withSupport Vector Machines[J], IEEE Transactions on Geoscience and Remote Sensing,2004,42(8):1778-1790.
[79] Dixon B and Candade N. Multispectral land use classification using neural networks andsupport vector machines: one or the other, or both?[J], International Journal of RemoteSensing,2008,29(4):1185-1206.
[80] Roli F and Fumera G. Support vector machines for remote-sensing image classification[C], inProc. EOS/SPIE Symposium,(Barcelona),2000.
[81] Hsu C W and Lin C J. A comparison of methods for multiclass support vector machines[J].IEEE Transactions on Neural Networks,2002,13(2):415-425.
[82] Platt J C, Cristianini N and Shawe-Taylor J. Large margin DAG's for multi-classclassification[C]. Advances in Neural Information Processing Systems,2000,12:547-553.
[83] Bruzzone L, Mingmin C and Marconcini M. A novel transductive SVM for semisupervisedclassification of remote-sensing images[J]. IEEE Transactions on Geoscience and RemoteSensing,2006,44(11):3363-3373.
[84] Demir B and Erturk S. Hyperspectral image classification using relevance vector machines[J].IEEE Transactions on Geoscience and Remote Sensing,2007,4(4):586-590.
[85] Foody G M. RVM-based multi-class classification of remotely sensed data[J]. InternationalJournal of Remote Sensing,2008,29(6):1817-1823.
[86] Zhang R and Ma J. An improved SVM method P-SVM for classification of remotely senseddata[J]. International Journal of Remote Sensing,2008,29(20):6029-6036.
[87] Quinlan J. Induction of decision trees[J]. Machine learning,1986,1(1):81-106.
[88] Friedl M and Brodley C. Decision tree classification of land cover from remotely senseddata[J]. Remote Sensing of Environment,1997,61(3):399-409.
[89] Friedl M, Brodley C and Strahler A. Maximizing land cover classification accuraciesproduced by decision trees at continental to global scales[J]. IEEE Transactions onGeoscience and Remote Sensing,1999,37(2):969-977.
[90] Hansen M, Dubayah R and Defries R. Classification trees: an alternative to traditional landcover classifiers[J], International Journal of Remote Sensing,1996,17(5):1075-1081.
[91] Belward A S and De Hoyos A. A comparison of supervised maximum likelihood and decisiontree classification for crop cover estimation from multitemporal LANDSAT MSS data[J],International Journal of Remote Sensing,1987,8(2):229-235.
[92] Wardlow B D and Egbert S L. Large-area crop mapping using time-series MODIS250mNDVI data: An assessment for the US Central Great Plains[J], Remote Sensing ofEnvironment,2008,112(3):1096-1116.
[93] Schneider A, McIver D K and Friedl M A, et al. Classification of urban areas at continentalscales using remotely sensed data[C], IEEE Geoscience and Remote Sensing Symposium,2001,5:2146-2148.
[94]吴成东,许可,韩中华,裴涛.基于粗糙集和决策树的数据挖掘方法[J].东北大学学报,2006,27(5):481-484.
[95]乔梅,韩文秀.基于Rough集的决策树算法[J].天津大学学报,2005,38(9):842-846.
[96] Asit Kumar Das and Jaya Sil. An efficient classifier design integrating rough set and setoriented database operations[J]. Applied Soft Computing,2011,11(2):2279-2285.
[97] Stefanov W L, Ramsey M S and Christensen P R. Monitoring urban land cover change: Anexpert system approach to land cover classification of semiarid to arid urban centers[J],Remote Sensing of Environment,2001,77(2):173-185
[98]贾海峰,刘雪华.环境遥感原理与应用[M],北京:清华大学出版社,2006:204.
[99]甘淑,袁希平,何大明.遥感专家分类系统在滇西北植被信息提取中的应用试验研究[J].云南大学学报(自然科学版),2003,25(6):553-557.
[100] Cohen, Y and Shoshany M. Integration of remote sensing, GIS and expert knowledge innational knowledge-based crop recognition in Mediterranean environment[J], InternationalJournal of Applied Earth Observation and Geoinformation,2002,4(1):75-87.
[101] Schmidt K S, Skidmore A K and Kloosterman E H, et al. Mapping coastal vegetation usingan expert system and hyperspectral imagery[J]. Photogrammetric Engineering and RemoteSensing,2004,70(6):703-715.
[102] Lucas R, Rowlands A and Brown A, et al. Rule-based classification of multi-temporalsatellite imagery for habitat and agricultural land cover mapping[J]. ISPRS Journal ofPhotogrammetry&Remote Sensing,2007,62(3):165-185.
[103] Zadeh L A. Fuzzy Sets. Information and Control[J].1965,8(3):338-353
[104] Liu Y M and Luo M K. Fuzzy Topology[M]. World Scientific,1997.
[105] Wu G and Zheng C. Fuzzy boundary and characteristic properties oforder-homomorphisms[J]. Fuzzy Sets and Systems1991,39(3):329-337.
[106] Tang X M, Kainz W and Fang Y. Reasoning about changes of land covers with fuzzysettings[J]. International Journal of Remote Sensing,2005,26(14):3025-3046.
[107] Andrew J T, Hugh G L and Peter M A, et al. Super-resolution target identification fromremotely sensed images using a hopfield neural network[J]. IEEE Transactions OnGeoscience and Remote Sensing,2001,39(4):781-796.
[108] Kittler J. Improving recognition rates by classifier combination: a theoretical Framework
[C]. Progress in Handwriting Recognition. Singapore: World Scientific Publishing,1997:231-248.
[109] Xu L, Krzyzak A and Suen C Y. Methods of combining multiple classifiers and theirapplications to handwriting recognition[J]. IEEE Transactions on Systems, Man andCybernetics,1992,22(3):418-435.
[110] Ho T K, Hull J J and Srihari S N. Decision combination in multiple classifier systems[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1994,16(1):66-75.
[111] Cho S B and Kim J H. Multiple network fusion using fuzzy logic[J]. IEEE Transactions onNeural Networks,1995,6(2):497-501.
[112] Lam L and Suen C Y. Optimal combination of pattern Classifiers[J]. Pattern RecognitionLetters,1995,6(9):945-954.
[113] Freund Y and Schapire R E. Experiments with a new Boosting algorithm. In: LorenzaSAITTA, ed. Machine Learning: Proceedings of the Thirteenth International Conference(ICML '96).1996, San Francisco, CA: Morgan Kaufmann:148-156.
[114] Breiman L. Bagging Predictors[J]. Machine Learning,1996,24(2):123-140.
[115] Lam L and Suen C Y. Application of majority voting to pattern recognition: An analysis ofits behaviour and performance[J]. IEEE Transactions on System, Man, and Cybernetics,Part A,1997,27(5):553-568.
[116] Kittler J, Hatef M and Duin R P, et al. On combining classifiers [J]. IEEE Transactions onPattern Analysis and Machine Intelligence,1998,20(3):226-239.
[117] Opitz D and Maclin R. Popular ensemble methods: an empirical study[J]. Journal ofArtificial Intelligence Research,1999,11(1):169-198.
[118] Jain A K, Duin R P W and Mao J-C. Statistical pattern recognition: A review[J]. IEEETransactions on Pattern Analysis and Machine Intelligence,2000,22(1):4-37.
[119] Skurichina M and Duin R P W. Bagging, Boosting and the random dubspace method forlinear classifiers [J]. Pattern Analysis and Applications,2002,5(2):121-135.
[120] Zhou Z H and Tang W. Selective ensemble of decision trees[J]. Lecture Notes in ComputerScience,2003,2639:476-483.
[121] Melville P and Mooney R J. Creating diversity in ensembles using artificial data[J].Information Fusion,2005,6(1):99-111.
[122]肖旭红,戴汝为.一种识别手写汉字的多分类器集成方法[J].自动化学报,1997,23(5):621-627.
[123]王岚,陈坷,迟惠生.多特征组合多分类器的方法用于“文本无关”的说话人辨认[J].北京大学学报(自然科学版),1998,34(2-3):143-I50.
[124] Zhou Z H and Yu Y. Adapt bagging to nearest neighbor classifiers[J]. Journal of ComputerScience and Technology,2005(1):48-54.
[125] Steele B M. Combining multiple classifiers: an application using spatial and remotelysensed information for land cover type mapping[J]. Remote Sensing of Environment,2000,74(3):545-556.
[126] Huang Z and Lees B G. Combining Non-Parametric Models for Multisource PredictiveForest Mapping[J]. Photogrammetric Engineering and Remote Sensing,2004,70(4):415-425.
[127] Foody G M, Boyd D S and Sanchez-Hernandez C. Mapping a specific class with anensemble of classifiers[J]. International Journal of Remote Sensing,2007,28(7):1733-1746.
[128] Serra P and Pons X. Monitoring farmers' decisions on Mediterranean irrigated crops usingsatellite image time series[J]. International Journal of Remote Sensing,2008,29(8):2293-2316.
[129] Maulik U and Chakraborty D A. A robust multiple classifier system for pixel classificationof remote sensing images[J]. Fundamenta Informaticae,2010,101(4):286-304.
[130] Zhang Y. Optimisation of building detection in satellite images by combining multispectralclassification and texture filtering[J]. ISPRS Journal of Photogrammetry and RemoteSensing,1999,54(1):50-60.
[131] Zhang Y. Texture-integrated classification of urban treed areas in High-ResolutionColor-Infrared imagery[J]. Photogrammetric Engineering and Remote Sensing,2001,67(12):1359-1365
[132]吴昊.综合纹理特征的高光谱遥感图像分类方法[J].计算机工程与设计,2012,33(5):1993-2006
[133]李厚强,刘政凯,林峰.基于分形理论的航空图像分类方法[J].遥感学报,2001,5(5):353-357.
[134] Benediktsson J A, Palmason J A and Sveinsson J R. Classification of hyperspectral datafrom urban areas based on extended morphological profiles[J]. IEEE Transactions onGeoscience and Remote Sensing,2005,43(3):480-491.
[135] Fauvel, M, Benediktsson J A and Chanussot J, et al. Spectral and spatial classification ofhyperspectral data using SVMs and morphological profiles[J]. IEEE Transactions onGeoscience and Remote Sensing,2005,43(3):3804-3815.
[136]汪闽,骆剑承,周成虎等.结合高斯马尔可夫随机场纹理模型与支撑向量机在高分辨率遥感图像上提取道路网[J].遥感学报,2005,9(3):271-276.
[137] Zhao Y D, Zhang L P and Li P X, et al. classification of high spatial resolution imageryusing improved gaussian markov random-field-based texture features[J]. IEEE Transactionson Geoscience and Remote Sensing,2007,45(5):1458-1468.
[138] Segl K, Reossner S and Heiden U, et al. Fusion of spectral and shape features foridentification of urban surface cover types using reflective and thermal hyperspectral data[J].ISPRS Journal of Photogrammetry and Remote Sensing,2003,58(1-2):99-112.
[139] Shackelford A and Davis X H. Hierarchical fuzzy classification approach for high-resolutionmultispectral data over urban areas[J]. IEEE Transactions on Geoscience and RemoteSensing,2003,41(9):1920-1932.
[140] Zhang L P, Huang X and Huang B, et al. A pixel shape index coupled with spectralinformation for classification of high spatial resolution remotely sensed imagery[J]. IEEETransactions on Geoscience and Remote Sensing,2006,44(10):2950-2961.
[141]黄昕,张良培,李平湘.融合形状和光谱的高空间分辨率遥感影像分类[J].遥感学报,2007,11(2):193-200.
[142] Huang X Zhang L P and Li P X, et al. classification and extraction of spatial features inurban areas using high-resolution multispectral imagery[J]. IEEE Geoscience and RemoteSensing Letters,2007,4(2):260-264.
[143] Huang X Zhang L P and Li P X. Classification of very high spatial resolution imagery basedon the fusion of edge and multispectral information[J]. Photogrammetric Engineering andRemote Sensing,2008,74(12):1585-1596.
[144] Congalton R G and Mead R A. A quantitative method to test for consistency and correctnessin photointerpretation[J]. Photogrammetric Engineering and Remote Sensing,1983,49(1):69-74.
[145] Aronoff S. The minimum accuracy value as an index of classification accuracy[J].Photogrammetric Engineering and Remote Sensing,1985,51(1):99-111.
[146] Koukoulas S and Blackburn G A. Introducing new indices for accuracy evaluation ofclassified images represnting semi-natural woodland environments[J]. PhotogrammetricEngineering and Remote Sensing,2001,67(4):499-510.
[147] Woodcock C E and Gopal S. Fuzzy set theory and thematic maps: accuracy assessment andarea estimation[J]. International Journal of Geographical Information Science,2000,14(2):153-172.
[148] Congalton R G, Oderwald R G and Mead R A. Assessing landsat classification accuracyusing discrete multivariate statistical techniques[J]. Photogrammetric Engineering andRemote Sensing,1983,49(12):1671-1678.
[149] Canters F. Evaluating the uncertainty of area estimates derived from fuzzy land-coverclassification[J]. Photogrammetric Engineering and Remote Sensing,1997,63(4):403–414.
[150] Foody G M. Status of land cover classification accuracy assessment[J]. Remote Sensing ofEnvironment,2002,80(1):185-201.
[151]柏延臣,王劲峰.遥感数据专题分类不确定性评价研究:进展、问题与展望[J].地球科学进展,2005,20(11):1218-1225.
[152] Foody G M, Campbell N A and Wood T F. Derivation and applications of probabilisticmeasures of class membership from the maximum likelihood classification[J].Photogrammetric Engineering and Remote Sensing,1992,58(9):1335-1341.
[153] Stephen V S and Raymond L C. Design and analysis for thematic map accuracy assessment:fundamental principles[J]. Remote Sensing of Environment,1998,64(3):331-344.
[154] Stephen V S. Sampling designs for assessing map accuracy[C]. Proceedings of the8thInternational Symposium on Spatial Accuracy Assessment in Natural Resources andEnvironmental Sciences, Shanghai, P. R. China,2008:8-15.
[155] Ge Y, li S P and Lakhan V C, et al. Exploring uncertainty in remotely sensed data withparallel coordinate plots[J]. International Journal of Applied Earth Observation andGeoinformation,2009,11(6):413-422.
[156] Rosenfield G H. Analysis of variance of thematic mapping experiment data[J],Photogrammetric Engineering and Remote Sensing,1981,47(12):1685-1692.
[157] Goodchild M F, Sun G Q and Yang S R. Development and test of and error model forcategorical data[J]. International Journal of Geographic Information Systems,1992,6(2):87-104.
[158] Congalton R G. A review of assessing the accuracy of classifications of remotely senseddata[J]. Remote Sensing of Environment,1991,37(1):35-46.
[159] Stevens Don L J R. Variable density grid-based sampling designs for continuous spatialpopulations[J]. Environmetrics,1997,8(3):164-195.
[160] Foody G M and Mathur A. The use of small training sets containing mixed pixels foraccurate hard image classification: Training on mixed spectral responses for classificationby a SVM [J]. Remote Sensing of Environment,2006,103(2):179-189.
[161] Li Z L and Huang P Z. Quantitative measures for spatial information of maps[J].International Journal of Geographical Information Science,2002,16(7):699-709.
[162] Cortes C and Vapnik V N. Support vector networks[J]. Machine Learning,1995,20:273-297.
[163] Boser B E, Guyon I M and Vapnik V N. A training algorithm for optimal marginclassifiers[C]. In Proc. Fifth Ann. Workshop on Comp. Learn Theory., Pittsburgh, PA,1992:144-152.
[164] Hsu C W and Lin C J. A comparison of methods for multi-class support vector machines[J].IEEE Transactions on Neural Networks,2002,13(2):415-425.
[165] Platt J C. Probabilities for Support Vector Machines[M]. Massachusetts: MIT Press,2000:61-74.
[166] Wu T F, Lin CH J and Weng R C. Probability estimates for multi-class classification bypairwise coupling[J]. Journal of Machine Learning Research,2004,5:975-1005.
[167] Chang C L. Fuzzy topological spaces[J]. Journal of Mathematical Analysis and Applications,1968,24(2):182-190.
[168] Liu K F and Shi W Z. Computation of fuzzy topological relations of spatial objects based oninduced fuzzy topology[J], International Journal of Geographical Information Systems,2006,20(8):857-883.
[169] Shi W Z and Liu K F. A fuzzy topology for computing the interior, boundary, and exterior ofspatial objects quantitatively in GIS[J]. Computers&Geosciences,2007,33(7):898-915.
[170] Shi W Z, Liu K F and Huang C Q. A fuzzy Topology-integrated area object extractionmethod[J]. IEEE Transactions on Geoscience and Remote Sensing,2010,48(1):147-154.
[171] Hsu C W, Chang C C and Lin C J. A practical guide to support vector classification,2010.
[Online]. Available:
[172] LekasK B, Likask A and Ga latsanos N P. A spatially constrained mixture model for imagesegmentation [J]. IEEE Transactions on Neural Networks,2005,16(2):494-498.
[173] Bezdek J C. Pattern recognition with fuzzy objective function algorithms [M]. New York:Plenum Press,1981:3-5.
[174]王顺凤，张建伟.基于Gibbs场与模糊C-均值聚类的脑MR图像分割[J].计算机应用,2008,28(7):1750-1752.
[175] LI S Z. Markov random field modeling in image analysis (Third Edition)[M]. Berlin:Springer-Verlag,2009:65-67.
[176] GC, C., J. AK. Markov random field texture models[J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,1983,1(1):25-39.
[177]杨龙平,丁宣浩,孙建明.一种新的基于小波变换的图像消噪方法[J].云南民族大学学报:自然科学版,2005,14(1):71-74.
[178] Suzuki T, Sugiyama M and Kanamori T, et al. Mutual information estimation reveals globalassociations between stimuli and biological processes[J]. BMC Bioinformatics,2009,10(1):1391-1445.
[179] Chang C C and Lin C J. LIBSVM: A Library for Support Vector Machines,2001.[Online].Available: http://www.csie.ntu. edu.tw/~cjlin/libsvm/.
[180] Theobald, D.M. and J.B. Norman.2006. Spatially-balanced sampling using The ReversedRandomized Quadrant-Recursive Raster algorithm: A User’s Guide for the RRQRR ArcGISv9.1tool. URL: http://www.nrel.colostate.edu/projects/starmap.
[181] Biernacki C. Initializing EM using the properties of its trajectories in Gaussian Mixtures[J].Statistics and Computing,2004,14(3):267-279.
[182] Fraley C. Algorithms for model-based Gaussian hierarchical clustering[J]. SIAM Journal onScientific Computing archive,1998,20(1):270–281.
[183] Fraley C and Raftery A E. Model-based clustering, discriminant analysis and densityestimation[J]. Journal of the American Statistical Association,2002,97(458):611–631.