遥感图像分类方法的研究
|
摘要
90年代卫星遥感在全球和区域尺度土地覆盖研究与应用方面取得了突破性进展，土地利用／覆盖遥感研究的新方法不断出现。遥感图像的分类是遥感数据在土地资源分析及应用的第一步，如何解决多类别图像的识别并满足一定的精度，是遥感图像研究中的一个关键问题，具有十分重要的意义。
与传统统计方法的分类器相比较，人工神经网络法不需要预先假设样本空间的参数化统计分布，正在被越来越普遍的应用于遥感图像分类的研究。在神经网络的不同算法中，应用和研究最多的是反向传播人工神经网络模型(简称BP网络模型)。
反向传播神经网络方法可用于遥感图像分类。本文在对BP网络分析的基础上，为提高网络的收敛速度和性能，提出了自己的一些建议，通过设置并改变样本的训练强度加快了网络学习的速度。另外还构造了一个用于混合像素分类的神经网络，输出层节点为各典型地物类别所占的百分比。分类前对数据进行主成分分析，实现特征提取的目的。在实际分类时仅使用波谱信息有时是不够的，可以添加一些辅助信息，如高程、植被指数等进行基于知识的推理判断。使用这些方法对实验区的TM图像进行了分类实验研究，取得了较好的效果。
At present, there is a far-reaching use of remote sensing (RS) imagine in the field of land use/land cover change (LUCC) research. The research on LUCC is a core for studies on the global changes. Also the research of spatial-temporal features of land use/land cover change is significantly important for better understanding land use/land cover change and environmental management for sustainable development. The techniques of classification are very important for land use/land cover change (LUCC),so LUCC study are on the basis of the image processing and classification system in the first. How to improve the accuracy of RS interpretation in order to promote the utility of RS technology is a urgent problem in RS application.
Compared with conventional statistic classifier, the artificial neural network (ANN) has been developed and applied to remote sensing data classification problem, which doesn't need suppose parameterized distribution of sample space in advance. ANN has complicated mapping capability. The back propagation neural network modal (BP model) is often been used.
Back propagation neural network classifier can solve the problems existing in the traditional classifiers and has been gradually used in the classification of remote sensing image. In order to accelerate the training speed, an improved BP method is to be presented after analyzing the BP model. Through setting a training intensity, the training time is reduced. As to the mix pixels, we construct a BP neural network which the nodes of input layer are the bands of remote sensing and the nodes of output layer are percent of several kinds of object. Before classification, Karhunen-Loeve transform is carried through for features abstract. In addition the spectrum information, many ancillary geo-information such as NDVI and Dem is contributed to judge. Synthesizing these methods, we have the test. Compared with classification of MLC method, the results show the improved method has not only the highest accuracy but also the fastest speed of classification.
与传统统计方法的分类器相比较，人工神经网络法不需要预先假设样本空间的参数化统计分布，正在被越来越普遍的应用于遥感图像分类的研究。在神经网络的不同算法中，应用和研究最多的是反向传播人工神经网络模型(简称BP网络模型)。
反向传播神经网络方法可用于遥感图像分类。本文在对BP网络分析的基础上，为提高网络的收敛速度和性能，提出了自己的一些建议，通过设置并改变样本的训练强度加快了网络学习的速度。另外还构造了一个用于混合像素分类的神经网络，输出层节点为各典型地物类别所占的百分比。分类前对数据进行主成分分析，实现特征提取的目的。在实际分类时仅使用波谱信息有时是不够的，可以添加一些辅助信息，如高程、植被指数等进行基于知识的推理判断。使用这些方法对实验区的TM图像进行了分类实验研究，取得了较好的效果。
At present, there is a far-reaching use of remote sensing (RS) imagine in the field of land use/land cover change (LUCC) research. The research on LUCC is a core for studies on the global changes. Also the research of spatial-temporal features of land use/land cover change is significantly important for better understanding land use/land cover change and environmental management for sustainable development. The techniques of classification are very important for land use/land cover change (LUCC),so LUCC study are on the basis of the image processing and classification system in the first. How to improve the accuracy of RS interpretation in order to promote the utility of RS technology is a urgent problem in RS application.
Compared with conventional statistic classifier, the artificial neural network (ANN) has been developed and applied to remote sensing data classification problem, which doesn't need suppose parameterized distribution of sample space in advance. ANN has complicated mapping capability. The back propagation neural network modal (BP model) is often been used.
Back propagation neural network classifier can solve the problems existing in the traditional classifiers and has been gradually used in the classification of remote sensing image. In order to accelerate the training speed, an improved BP method is to be presented after analyzing the BP model. Through setting a training intensity, the training time is reduced. As to the mix pixels, we construct a BP neural network which the nodes of input layer are the bands of remote sensing and the nodes of output layer are percent of several kinds of object. Before classification, Karhunen-Loeve transform is carried through for features abstract. In addition the spectrum information, many ancillary geo-information such as NDVI and Dem is contributed to judge. Synthesizing these methods, we have the test. Compared with classification of MLC method, the results show the improved method has not only the highest accuracy but also the fastest speed of classification.
引文
1．梅安新．遥感导论[M]，北京：高等教育出版社，2001，1-3
2．张永生．遥感图象信息系统[M]，北京：科学出版社，2000，26-33
3．朱亮璞．遥感地质学[M]，北京：地质出版社，1999，14-19
4．边肇祺，张学工．模式识别[M]，北京：清华大学出版社，2000，176-228，47-55
5．吴简彤，王建华．神经网络技术及其应用[M]，哈尔滨：哈尔滨工程大学出版社，1998，10-20
6．杨立民，朱智良．全球及区域尺度土地覆盖土地利用遥感研究的现状和展望[J]，自然资源学报．1999．14(4)：340-344
7．Martin T. Hagan, Howard B. Demuth, Mark H. Beale著．戴葵等译．神经网络设计[M]，北京：机械工业出版社，2000，34-47
8．蒋宗礼．人工神经网络导论[M]，北京：高等教育出版社，2001，39-52
9．詹庆明，肖映辉．城市遥感技术[M]，武汉：武汉测绘科技大学出版社，1999
10．焦李成．神经网络的应用与实现[M]，西安：西安电子科技大学出版社，1996
11．黄德双．神经网络模式识别系统理论[M]，北京：电子工业出版社，1996
12．李朝峰，王桂梁．模糊控制BP网络的遥感图象分类方法研究[J]，中国矿业大学学报(自然科学版)，2001年03期
13．熊桢，郑兰芬，童庆禧．分层神经网络分类算法[J]，测绘学报，2000：29(3)229～234
14．于秀兰，钱国蕙等．多光谱遥感图像BP网分类器学习样本选取法的研究[J]
15．杜惠茜，梅文博，李德生．一种改进的BP神经网络在遥感图像分类中的应用[J]，北京理工大学学报，1998年04期
16．贾永红，张春森，王爱平．基于BP神经网络的多源遥感影像分类[J]．
17．李厚强，王宜主，刘政凯．一种用于多类别遥感图像分类的新方法，遥感学报，1997，1(4)：257-261
18．金忠，杨静宇，胡钟山．有效最佳鉴别特征的抽取与维数问题．计算机学报 2000，23(1)：108-112
19．朱述龙．快速近似主成分分析算法[J]，遥感学报，1999(3)：43-47
20．杨存建，周成虎，TM 影像的居民地信息提取方法研究，遥感学报，2000(4)：146-150
21．朱小鸽，多重主成分分析及在地质构造信息提取中的应用，遥感学报，2000(4)：299-303
22．秦其明，陆荣建．分形与神经网络方法在卫星数字图像分类中的应用[J]，北京
大学学报(自然科学版)，2000(36)：858-864
23．刘建平，赵英时，孙淑玲．高光谱遥感数据最佳波段选择方法试验研究[J]，遥感技术与应用，2001(16)：7-13
24．王耀南．卫星遥感图像的神经网络自动识别分类[J]，湖南大学学报 1998(4)：61-66
25．朱明，王俊普．一种聚类学习的新方法[J]，模式识别与人工智能，2000(3)：262-265。
26．张铃，张钹，殷海风．多层前向网络的交叉覆盖设计算法[J]，软件学报，1999(7)：738-742
27．李本纲，陶澍．一种利用多时相TM影像分析地表植被变化的新方法[J]，遥感学报，2000，4(4)：295～298
28．吴健平，杨星卫．遥感监督分类中训练样本的纯化[J]，国土资源遥感，1996(1)：36-41
29．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(3)233～239
30．邸凯昌，李德仁，李德毅．基于空间数据发掘的遥感图像分类方法研究[J]，武汉测绘科技大学学报，2000：1．25(1)42-48
31．苑进，基于RBF神经网络的遥感图像分类和基于结构特征的水体识别[D]，山东泰安：山东科技大学，2002
32. L. Bruzzone, F. Roli, and S. B. Serpico, "Structured Neural Networks for Signal Classification", Signal Processing, 64, 1998, pp. 271-290
33. L. Bruzzone, D. Fernandez Prieto, An incremental-learning neural network for the classifications of remote-sensing images[J], Pattern Recognition Letters 20(1999) 1241-1248
34. H. Baltzakis, N. Papamarkos, A new signature verifcation technique based on a two-stage neural network classifer, Engineering Applications of Artifcial Intelligence 14(2001): 95-103
35. Yee Leung. Intelligent spatial decision support systems. In: Advances in Spatial Science. Germany: Springer, 1997.
36. Atkinson P M and Tatnall A R L. Neural networks in remote sensing. INT J Remote Sensing, 1997, 18(4): 699-709.
37. Benediktsson J A, Swain P H, Ersoy O K. Neural network approaches versus statistical methods in classification of multi-source remote sensing data. IEEE Transactions on Geo-science and Remote Sensing, 1990, 28: 540-552.
38. Bischof H, Schneider W, Pinz A J. Multi-spectral classification of landsat images using neural networks. IEEE Transactions on Geo-science and Remote Sensing, 1992,
30: 482-490.
39. Paola J D, Schowengerdt R A. A review and analysis of back-propagation neural networks for classification of remotely-sen-sed multi-spectral imagery, INT. J. REMOTE SENSING, 1995, 16; 3033-3058.
40. D. S. Broomhead and D. Low, Multivariable functional interpolattion and adaptive networks, Complex Systems, vol.2, 1988, pp. 321-355.
41. Serpico, S. B., Bruzzone, L., Roli, F., 1996. An experimental comparison of neural and statistical non-parametric algorithms for supervised classification of remote sensing images. Pattern Recognition Letters 17(13), 1331～1341.
42. H. Baltzakis, N. Papamarkos, A new signature verification technique based on a two-stage neural network classifier Engineering Applications of Artificial Intelligence 14(2001)95-103
43. Shlomo Geva, Kurt Malmstrom, Joaquin Sitte. Local cluster neural net: Architecture, training and applications. Neurocomputing 20(1998)35～56
44. Dominique Durand, Jerome Bijaoui, Francois Cauneau, Optical Remote Sensing of Shallow-Water Environmental Parameters: A Feasibility Study, REMOTE SENS. ENVIRON. 73: 152-161(2000)
45. Qiang Ji, Mauro S. Costa, Robert M. Haralick, Arobust linear least -squares estimation of camera exterior orientation using multiple geometric features. Journal of Photogrammetry & Remote Sensing 55(2000) 75-93
46. R. Bernstein, V. Di Gesu, A combined analysis to extract objects in remote sensing images. Pattern Recognition Letters 20(1999) 1407～1414
47. Arcot Sowmya, John Trinder, Modelling and representation issues in automated feature extraction from aerial and satellite images. Journal of Photogrammetry & Remote Sensing 55(2000) 34-47
48. Younian Wang, Principles and applications of structural image matching. Journal of Photogrammetry & Remote Sensing 53(1998)154-165
49. Giles M. Foody, Estimation of sub-pixel land cover composition in the presence of untrained classes Computers & Geosciences 26(2000)469-478
50. V. F. Kxapivin and G. W. Phillips, A Remote Sensing-Based Expert System to Study the Aral-Caspian Aquageosystem Water Regime. REMOTE SENS. ENVIRON. 75: 201-215(2001)
2．张永生．遥感图象信息系统[M]，北京：科学出版社，2000，26-33
3．朱亮璞．遥感地质学[M]，北京：地质出版社，1999，14-19
4．边肇祺，张学工．模式识别[M]，北京：清华大学出版社，2000，176-228，47-55
5．吴简彤，王建华．神经网络技术及其应用[M]，哈尔滨：哈尔滨工程大学出版社，1998，10-20
6．杨立民，朱智良．全球及区域尺度土地覆盖土地利用遥感研究的现状和展望[J]，自然资源学报．1999．14(4)：340-344
7．Martin T. Hagan, Howard B. Demuth, Mark H. Beale著．戴葵等译．神经网络设计[M]，北京：机械工业出版社，2000，34-47
8．蒋宗礼．人工神经网络导论[M]，北京：高等教育出版社，2001，39-52
9．詹庆明，肖映辉．城市遥感技术[M]，武汉：武汉测绘科技大学出版社，1999
10．焦李成．神经网络的应用与实现[M]，西安：西安电子科技大学出版社，1996
11．黄德双．神经网络模式识别系统理论[M]，北京：电子工业出版社，1996
12．李朝峰，王桂梁．模糊控制BP网络的遥感图象分类方法研究[J]，中国矿业大学学报(自然科学版)，2001年03期
13．熊桢，郑兰芬，童庆禧．分层神经网络分类算法[J]，测绘学报，2000：29(3)229～234
14．于秀兰，钱国蕙等．多光谱遥感图像BP网分类器学习样本选取法的研究[J]
15．杜惠茜，梅文博，李德生．一种改进的BP神经网络在遥感图像分类中的应用[J]，北京理工大学学报，1998年04期
16．贾永红，张春森，王爱平．基于BP神经网络的多源遥感影像分类[J]．
17．李厚强，王宜主，刘政凯．一种用于多类别遥感图像分类的新方法，遥感学报，1997，1(4)：257-261
18．金忠，杨静宇，胡钟山．有效最佳鉴别特征的抽取与维数问题．计算机学报 2000，23(1)：108-112
19．朱述龙．快速近似主成分分析算法[J]，遥感学报，1999(3)：43-47
20．杨存建，周成虎，TM 影像的居民地信息提取方法研究，遥感学报，2000(4)：146-150
21．朱小鸽，多重主成分分析及在地质构造信息提取中的应用，遥感学报，2000(4)：299-303
22．秦其明，陆荣建．分形与神经网络方法在卫星数字图像分类中的应用[J]，北京
大学学报(自然科学版)，2000(36)：858-864
23．刘建平，赵英时，孙淑玲．高光谱遥感数据最佳波段选择方法试验研究[J]，遥感技术与应用，2001(16)：7-13
24．王耀南．卫星遥感图像的神经网络自动识别分类[J]，湖南大学学报 1998(4)：61-66
25．朱明，王俊普．一种聚类学习的新方法[J]，模式识别与人工智能，2000(3)：262-265。
26．张铃，张钹，殷海风．多层前向网络的交叉覆盖设计算法[J]，软件学报，1999(7)：738-742
27．李本纲，陶澍．一种利用多时相TM影像分析地表植被变化的新方法[J]，遥感学报，2000，4(4)：295～298
28．吴健平，杨星卫．遥感监督分类中训练样本的纯化[J]，国土资源遥感，1996(1)：36-41
29．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(3)233～239
30．邸凯昌，李德仁，李德毅．基于空间数据发掘的遥感图像分类方法研究[J]，武汉测绘科技大学学报，2000：1．25(1)42-48
31．苑进，基于RBF神经网络的遥感图像分类和基于结构特征的水体识别[D]，山东泰安：山东科技大学，2002
32. L. Bruzzone, F. Roli, and S. B. Serpico, "Structured Neural Networks for Signal Classification", Signal Processing, 64, 1998, pp. 271-290
33. L. Bruzzone, D. Fernandez Prieto, An incremental-learning neural network for the classifications of remote-sensing images[J], Pattern Recognition Letters 20(1999) 1241-1248
34. H. Baltzakis, N. Papamarkos, A new signature verifcation technique based on a two-stage neural network classifer, Engineering Applications of Artifcial Intelligence 14(2001): 95-103
35. Yee Leung. Intelligent spatial decision support systems. In: Advances in Spatial Science. Germany: Springer, 1997.
36. Atkinson P M and Tatnall A R L. Neural networks in remote sensing. INT J Remote Sensing, 1997, 18(4): 699-709.
37. Benediktsson J A, Swain P H, Ersoy O K. Neural network approaches versus statistical methods in classification of multi-source remote sensing data. IEEE Transactions on Geo-science and Remote Sensing, 1990, 28: 540-552.
38. Bischof H, Schneider W, Pinz A J. Multi-spectral classification of landsat images using neural networks. IEEE Transactions on Geo-science and Remote Sensing, 1992,
30: 482-490.
39. Paola J D, Schowengerdt R A. A review and analysis of back-propagation neural networks for classification of remotely-sen-sed multi-spectral imagery, INT. J. REMOTE SENSING, 1995, 16; 3033-3058.
40. D. S. Broomhead and D. Low, Multivariable functional interpolattion and adaptive networks, Complex Systems, vol.2, 1988, pp. 321-355.
41. Serpico, S. B., Bruzzone, L., Roli, F., 1996. An experimental comparison of neural and statistical non-parametric algorithms for supervised classification of remote sensing images. Pattern Recognition Letters 17(13), 1331～1341.
42. H. Baltzakis, N. Papamarkos, A new signature verification technique based on a two-stage neural network classifier Engineering Applications of Artificial Intelligence 14(2001)95-103
43. Shlomo Geva, Kurt Malmstrom, Joaquin Sitte. Local cluster neural net: Architecture, training and applications. Neurocomputing 20(1998)35～56
44. Dominique Durand, Jerome Bijaoui, Francois Cauneau, Optical Remote Sensing of Shallow-Water Environmental Parameters: A Feasibility Study, REMOTE SENS. ENVIRON. 73: 152-161(2000)
45. Qiang Ji, Mauro S. Costa, Robert M. Haralick, Arobust linear least -squares estimation of camera exterior orientation using multiple geometric features. Journal of Photogrammetry & Remote Sensing 55(2000) 75-93
46. R. Bernstein, V. Di Gesu, A combined analysis to extract objects in remote sensing images. Pattern Recognition Letters 20(1999) 1407～1414
47. Arcot Sowmya, John Trinder, Modelling and representation issues in automated feature extraction from aerial and satellite images. Journal of Photogrammetry & Remote Sensing 55(2000) 34-47
48. Younian Wang, Principles and applications of structural image matching. Journal of Photogrammetry & Remote Sensing 53(1998)154-165
49. Giles M. Foody, Estimation of sub-pixel land cover composition in the presence of untrained classes Computers & Geosciences 26(2000)469-478
50. V. F. Kxapivin and G. W. Phillips, A Remote Sensing-Based Expert System to Study the Aral-Caspian Aquageosystem Water Regime. REMOTE SENS. ENVIRON. 75: 201-215(2001)