基于主题模型的高分辨率遥感影像变化检测
|
摘要
随着科学技术的日益发展以及国计民生水平的不断提高,卫星遥感技术逐渐在各个方面发挥着不可替代的重要作用,而作为遥感技术的主要研究方向之一的遥感变化检测技术更是有着广泛而出色的应用,大到军事战争、灾难预警和救援,小到城市建设规划、农作物生长监控等,都可以看到它的身影。目前的遥感技术,正不断向着三高、三多的趋势发展,即高空间分辨率、高时间分辨率、高光谱分辨率以及多平台、多角度、多传感器,这也使得遥感影像数据量巨大,目标众多、地物信息繁杂的特点越来越明显,给传统的遥感变化检测方法带来了极大的挑战,迫切需要新的更加完善的变化检测方法的提出。
主题模型是一种最早起源于自然语言文本处理领域的概率生成模型,是在词包模型基础上的扩展,其主旨是通过对单词在文档层面的共现信息进行分析,抽象出文档中语义相关的隐藏主题信息。主题模型在数据降维,模糊搜索、目标分类等方面都有着广泛的应用。随着视觉单词概念的提出与研究的深入,主题模型也开始逐渐应用于数字图像处理领域并有着很好的效果。
本文提出一种基于主题模型的面向高分辨率遥感影像的变化检测方法,以前后两期遥感影像的像素点对为基本单位,提取相关度、均值、标准差、斜率、截距等低层次特征,并根据这些特征形成视觉单词特征,再通过经典的主题模型——潜在狄利克雷分配模型发掘出其隐含的变化与不变的主题信息,以达到遥感影像变化检测的目的,实验表明该方法能够有效地进行高分辨率遥感影像的变化检测。
With the development of science and technology as well as the continuous improvement of the level of people's livelihood,satellite remote sensing technology is gradually playing an irreplaceable role in many aspects,as one of the main research direction of it,remote sensing change detection technology has even more extensive and excellent applications. From military wars、disaster warning and relief to city construction planning and crop growth monitoring, it can appear in all these fields. Nowadays, remote sensing technology is constantly trending towards Three-High and Three-? Multi , which means high spatial resolution、high time resolution、high spectral resolution and multi-platform、multi-angle、multi-sensor. This trend makes the characteristics of remote sensing itself more and more obvious, namely huge amount of data、many objects and complex information of surface features. This also brings a great challenge to the traditional remote sensing change detection methods. Therefore, it is indeed urgent to propose new and more improved change detection methods.
Topic model is a probabilistic generative model, and it’s first orginated from the natural language text processing field. It is the extension of Bag of Words(BOW) model,aiming at abstracting the semantic-related latent topic information of the document by analysising the co-occurrence information of the words on the document level. Topic model has being widely used in the tasks such as dimensionality reduction of data、ambiguous searching and object categorization and so on. With the proposing of concept of visual-words and the advancing in research, topic model is now gradually applied in the field of digital image processing and has achieved well results.
A topic model based change detection method is proposed for high resolution remote sensing images in this paper. It takes the pixel pairs of bi-temporal remote sensing images as the basic unit, and extracts their low-level features, such as relevacy, mean value, standard deviation, slope and intercept. Then visual words on the basis of these features are generated. After that, the classical topic model of latent dirichlet allocation is utilized to find the latent topic information, that is, changed topic or unchanged topic, thereby the goal of change detection will be achieved. Experiments show that this method could effectively detect changes in high resolution remote sensing images.
主题模型是一种最早起源于自然语言文本处理领域的概率生成模型,是在词包模型基础上的扩展,其主旨是通过对单词在文档层面的共现信息进行分析,抽象出文档中语义相关的隐藏主题信息。主题模型在数据降维,模糊搜索、目标分类等方面都有着广泛的应用。随着视觉单词概念的提出与研究的深入,主题模型也开始逐渐应用于数字图像处理领域并有着很好的效果。
本文提出一种基于主题模型的面向高分辨率遥感影像的变化检测方法,以前后两期遥感影像的像素点对为基本单位,提取相关度、均值、标准差、斜率、截距等低层次特征,并根据这些特征形成视觉单词特征,再通过经典的主题模型——潜在狄利克雷分配模型发掘出其隐含的变化与不变的主题信息,以达到遥感影像变化检测的目的,实验表明该方法能够有效地进行高分辨率遥感影像的变化检测。
With the development of science and technology as well as the continuous improvement of the level of people's livelihood,satellite remote sensing technology is gradually playing an irreplaceable role in many aspects,as one of the main research direction of it,remote sensing change detection technology has even more extensive and excellent applications. From military wars、disaster warning and relief to city construction planning and crop growth monitoring, it can appear in all these fields. Nowadays, remote sensing technology is constantly trending towards Three-High and Three-? Multi , which means high spatial resolution、high time resolution、high spectral resolution and multi-platform、multi-angle、multi-sensor. This trend makes the characteristics of remote sensing itself more and more obvious, namely huge amount of data、many objects and complex information of surface features. This also brings a great challenge to the traditional remote sensing change detection methods. Therefore, it is indeed urgent to propose new and more improved change detection methods.
Topic model is a probabilistic generative model, and it’s first orginated from the natural language text processing field. It is the extension of Bag of Words(BOW) model,aiming at abstracting the semantic-related latent topic information of the document by analysising the co-occurrence information of the words on the document level. Topic model has being widely used in the tasks such as dimensionality reduction of data、ambiguous searching and object categorization and so on. With the proposing of concept of visual-words and the advancing in research, topic model is now gradually applied in the field of digital image processing and has achieved well results.
A topic model based change detection method is proposed for high resolution remote sensing images in this paper. It takes the pixel pairs of bi-temporal remote sensing images as the basic unit, and extracts their low-level features, such as relevacy, mean value, standard deviation, slope and intercept. Then visual words on the basis of these features are generated. After that, the classical topic model of latent dirichlet allocation is utilized to find the latent topic information, that is, changed topic or unchanged topic, thereby the goal of change detection will be achieved. Experiments show that this method could effectively detect changes in high resolution remote sensing images.
引文
[1]常庆瑞.遥感技术导论[M].科学出版社, 2004
[2]李德仁.论21世纪遥感与GIS的发展[J].武汉大学学报.信息科学版, 2003, 28 (2): 127-131
[3]马建文等.遥感变化检测技术发展综述[J].地球科学进展, 2004(4) 192-196
[4]张振龙,曾志远,李硕,胡子付.遥感变化检测方法研究综述[J].遥感信息,2005,5:64-66.
[5] Blei D M, Lafferty J D. Topic Models[M]. Taylor and Francis, 2009
[6] Salton, G. and Buckley, C. Term-weighting approaches in automatic text retrieval[J]. Information Processing & Management, 1988,24(5): 513–523.
[7] Deerwester, S., Dumais, S.T., etc. Indexing By Latent Semantic Analysis[J]. Journal of the American Society For Information Science, 1990,41:391-407.
[8] T.Hofmann. Probabilistic latent semantic indexing[J]. Procecdings of the Twenty-second Annual international SIGIR Conference on Research and Development in Information Retrievel, 1999,99:50-57.
[9] Blei D M, Ng A Y, Jordan M I. Latent Dirichlet allocation[J]. Journal of Machine Learning Research, 2003, 3 : 993-1022.
[10] D. Blei and J. Lafferty. A Correlated Topic Model of Science[J]. The Annals of Applied Statistics, 2007(1):17-35
[11] D. Blei and J. Lafferty. Dynamic topic models[J]. In Proceedings of the 23rd International Conference on Machine Learning, 2006
[12] D. Blei, J. McAuliffe. Supervised topic models[J]. Neural Information Processing Systems , 2007,21.
[13]钟家强.基于多时相遥感图像的变化检测[J].国防科技大学博士论文. 2005
[14]赵英时.遥感应用分析原理与方法[M].科学出版社. 2003
[15] Jensen J.R.. Introductory digital image processing: a remote sensing perspective Englewood Clifs:Prentice Hall. 1996
[16]方圣辉,佃袁勇,李微.基于边缘特征的变化检测方法研究[J].武汉大学学报(信息科学版), 2005, 30(2): 136-138.
[17]张凤玉,遥感图像变化检测方法研究[J],西安电子科技大学硕士论文, 2010
[18]佃袁勇,基于遥感影像的变化检测研究[J],武汉大学硕士论文, 2005
[19]孙晓霞,张继贤,燕琴,高井祥.遥感影像变化检测方法综述及展望[J],遥感信息, 2011.1:119-123
[20]吴磊.视觉语言分析:从底层视觉特征表达到语义距离学习[J].中国科学技术大学博士论文. 2010.
[21] Mohsilovic A,Gomes J,Rogowitz B. Isee: Perceptual features for image library navigation[J]. SPIE: Human vision and electronic imaging, 2002, 4662: 266-277
[22] Fredembach C,Schroder M,Susstrunk S. Eigenregions for image classification[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004,26(12): 1645-1649
[23] Marti J,Freixenet J,Batlle J,Casals A. A new approach to outdoor scene description based on learning and top-down segmentation[J]. Image and Vision Computing, 2001,19(14): 1041-1055
[24] Aksoy S,Koperski K,Tusk C,Marchisio G,Tilton J C. Learning Bayesian classifiers for scene classification with a visual grammar[J]. IEEE Transactions on Geoscience and Remote Sensing,2005,43(3): 581-589
[25] Fischler M A,Elschlager R A. The representation and matching of pictorial structures[J]. IEEE Transactions on Computer, 1973,22(1): 67-92
[26] Zhu L,Rao A B,Zhang A D. Theory of keyblock-based image retrieval[J]. Acm Transactions on Information Systems, 2002,20(2): 224-257
[27] Fergus, R., Fei-Fei, L., Perona, P., etc. Learning object categories from google's image search [A]. In Internation Conference on Computer Vision[C], 2005; 1816-1823.
[28] Fergus R,Perona P,Zisserman A. Object class recognition by unsupervised scale-invariant learning[J]. IEEE COmputer Society Conference on Computer Vision and Pattern Recognition, 2005b,2: 264-271
[29] Csurka, G., Dance, C., Fan, L., etc. visual categorization with bags of keypoints [A]. In Proc. European Conf. Computer Vision Workshop Statistical Learning in Computer Vision[C], 2004.
[30]Zhang H,Berg A,Maire M,Malik J. SVM-KNN: Discriminative nearest neighbor classification for visual category recognition[J]. IEEE Computer Society Conference on Computer VIsion and Pattern Recognition,2006, 2: 2126-2136
[31]Zhang J,Marszalek M,Lazebnik S,Schmid C. Local features and kernels for classification of texture and object categories: A comprehensive study[J]. International Journal of Computer Vision, 2007,73(2): 213-238
[32] Sebastiani, F. Machine learning in automated text categorization [J]. Acm Computing Surveys, Mar, 2002, 34 (1): 1-47.
[33] Leung, T., Malik, J. Representing and recognizing the visual appearance of materials using three-dimensional textons [J]. International Journal of Computer Vision, 2001, 43 (1): 29-44.
[34] Sivic, J., Zisserman, A. Video Google: A text retrieval approach to object matching in videos [A]. In Proc. 9th IEEE Int'l Conf. Computer Vision[C], 2003; 1470-1477.
[35] Cula, O., Dana, K. Compact representation of bidirectional textural functions [A]. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition[C], 2001; 1041-1047.
[36] Vogel, J., Schiele, B. A semantic typicality measure for natural scene categorization [A]. In Proc. 26th DAGM Sympo. Pattern Recognition[C], 2004; 195-203.
[37] Sivic, J., Russell, B. C., Efros, A. A., etc. Discovering objects and their location in images [A]. In Tenth IEEE International Conference on Computer Vision, Vols 1 and 2, Proceedings[C], 2005; 370-377.
[38] Jurie, F., Triggs, B. Creating efficient codebooks for visual recognition [A]. In Proc. 10th IEEE Int'l Conf. Computer Vision[C], 2005; 604-610.
[39] Maree, R., Geurts, P., Piater, J., etc. Random subwindows for robust image classification [A]. In Proc. IEEE Conf. Computer Vision and Pattern Recognition[C], 2005; 34-40.
[40] Moosmann, F., Larlus, D., Jurie, F. Learning Saliency Maps for Object Categorization [A]. In Proc. European Conf. Computer Vision Workshop Representation and Use of Prior Knowlodge[C], 2006.
[41] Moosmann, F., Larlus, D., Jurie, F. Fast discriminative visual codebooks using randomized clustering forests [A]. In In Advances in Neural Information Processing Systems(Nips)[C], 2006; 985-992.
[42] Perronnin, F. Universal and adapted vocabularies for Generic Visual Categorization [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, Jul, 2008, 30 (7): 1243-1256.
[43] Yang, Y., Newsam, S. Bag-Of-Visual-Words and Spatial Extensions for Land-Use Classification [A]. In ACM SIGSPATIAL International Conference on Advances in Geographic InformationSystems[C], 2010.
[44] Jiang, Y. G., Ngo, C. W. Visual word proximity and linguistics for semantic video indexing and near-duplicate retrieval [J]. Computer Vision and Image Understanding 2009, 113 (3): 405-414.
[45] van Gemert, J. C., Veenman, C. J., Smeulders, A. W. M., etc. Visual Word Ambiguity [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, Jul, 2010, 32 (7): 1271-1283.
[46] Arthur Dempster, Nan Laird, and Donald Rubin. Maximum likelihood from incomplete data via the EM algorithm[J]. Journal of the Royal Statistical Society, Series B, 1977,39(1):1–38,
[47] Winn J M. Variational message passing and it's application[D]. Cambridge: Uinversity of Cambridge, 2003.
[48] Im J, Jensen J R. A change detection model based on neighborhood correlation image analysis and decision tree classification[J]. Remote Sensing of Enviornment, 2005, 99(3): 326-340.
[49] Xu Sheng, Fang Tao, Li Deren, Wang Shiwei. Object classification of aerial image with Bag-of-Visual-Words[J]. IEEE Geoscience and Remote Sensing Letters, 2010, 7(2): 366-370.
[2]李德仁.论21世纪遥感与GIS的发展[J].武汉大学学报.信息科学版, 2003, 28 (2): 127-131
[3]马建文等.遥感变化检测技术发展综述[J].地球科学进展, 2004(4) 192-196
[4]张振龙,曾志远,李硕,胡子付.遥感变化检测方法研究综述[J].遥感信息,2005,5:64-66.
[5] Blei D M, Lafferty J D. Topic Models[M]. Taylor and Francis, 2009
[6] Salton, G. and Buckley, C. Term-weighting approaches in automatic text retrieval[J]. Information Processing & Management, 1988,24(5): 513–523.
[7] Deerwester, S., Dumais, S.T., etc. Indexing By Latent Semantic Analysis[J]. Journal of the American Society For Information Science, 1990,41:391-407.
[8] T.Hofmann. Probabilistic latent semantic indexing[J]. Procecdings of the Twenty-second Annual international SIGIR Conference on Research and Development in Information Retrievel, 1999,99:50-57.
[9] Blei D M, Ng A Y, Jordan M I. Latent Dirichlet allocation[J]. Journal of Machine Learning Research, 2003, 3 : 993-1022.
[10] D. Blei and J. Lafferty. A Correlated Topic Model of Science[J]. The Annals of Applied Statistics, 2007(1):17-35
[11] D. Blei and J. Lafferty. Dynamic topic models[J]. In Proceedings of the 23rd International Conference on Machine Learning, 2006
[12] D. Blei, J. McAuliffe. Supervised topic models[J]. Neural Information Processing Systems , 2007,21.
[13]钟家强.基于多时相遥感图像的变化检测[J].国防科技大学博士论文. 2005
[14]赵英时.遥感应用分析原理与方法[M].科学出版社. 2003
[15] Jensen J.R.. Introductory digital image processing: a remote sensing perspective Englewood Clifs:Prentice Hall. 1996
[16]方圣辉,佃袁勇,李微.基于边缘特征的变化检测方法研究[J].武汉大学学报(信息科学版), 2005, 30(2): 136-138.
[17]张凤玉,遥感图像变化检测方法研究[J],西安电子科技大学硕士论文, 2010
[18]佃袁勇,基于遥感影像的变化检测研究[J],武汉大学硕士论文, 2005
[19]孙晓霞,张继贤,燕琴,高井祥.遥感影像变化检测方法综述及展望[J],遥感信息, 2011.1:119-123
[20]吴磊.视觉语言分析:从底层视觉特征表达到语义距离学习[J].中国科学技术大学博士论文. 2010.
[21] Mohsilovic A,Gomes J,Rogowitz B. Isee: Perceptual features for image library navigation[J]. SPIE: Human vision and electronic imaging, 2002, 4662: 266-277
[22] Fredembach C,Schroder M,Susstrunk S. Eigenregions for image classification[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004,26(12): 1645-1649
[23] Marti J,Freixenet J,Batlle J,Casals A. A new approach to outdoor scene description based on learning and top-down segmentation[J]. Image and Vision Computing, 2001,19(14): 1041-1055
[24] Aksoy S,Koperski K,Tusk C,Marchisio G,Tilton J C. Learning Bayesian classifiers for scene classification with a visual grammar[J]. IEEE Transactions on Geoscience and Remote Sensing,2005,43(3): 581-589
[25] Fischler M A,Elschlager R A. The representation and matching of pictorial structures[J]. IEEE Transactions on Computer, 1973,22(1): 67-92
[26] Zhu L,Rao A B,Zhang A D. Theory of keyblock-based image retrieval[J]. Acm Transactions on Information Systems, 2002,20(2): 224-257
[27] Fergus, R., Fei-Fei, L., Perona, P., etc. Learning object categories from google's image search [A]. In Internation Conference on Computer Vision[C], 2005; 1816-1823.
[28] Fergus R,Perona P,Zisserman A. Object class recognition by unsupervised scale-invariant learning[J]. IEEE COmputer Society Conference on Computer Vision and Pattern Recognition, 2005b,2: 264-271
[29] Csurka, G., Dance, C., Fan, L., etc. visual categorization with bags of keypoints [A]. In Proc. European Conf. Computer Vision Workshop Statistical Learning in Computer Vision[C], 2004.
[30]Zhang H,Berg A,Maire M,Malik J. SVM-KNN: Discriminative nearest neighbor classification for visual category recognition[J]. IEEE Computer Society Conference on Computer VIsion and Pattern Recognition,2006, 2: 2126-2136
[31]Zhang J,Marszalek M,Lazebnik S,Schmid C. Local features and kernels for classification of texture and object categories: A comprehensive study[J]. International Journal of Computer Vision, 2007,73(2): 213-238
[32] Sebastiani, F. Machine learning in automated text categorization [J]. Acm Computing Surveys, Mar, 2002, 34 (1): 1-47.
[33] Leung, T., Malik, J. Representing and recognizing the visual appearance of materials using three-dimensional textons [J]. International Journal of Computer Vision, 2001, 43 (1): 29-44.
[34] Sivic, J., Zisserman, A. Video Google: A text retrieval approach to object matching in videos [A]. In Proc. 9th IEEE Int'l Conf. Computer Vision[C], 2003; 1470-1477.
[35] Cula, O., Dana, K. Compact representation of bidirectional textural functions [A]. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition[C], 2001; 1041-1047.
[36] Vogel, J., Schiele, B. A semantic typicality measure for natural scene categorization [A]. In Proc. 26th DAGM Sympo. Pattern Recognition[C], 2004; 195-203.
[37] Sivic, J., Russell, B. C., Efros, A. A., etc. Discovering objects and their location in images [A]. In Tenth IEEE International Conference on Computer Vision, Vols 1 and 2, Proceedings[C], 2005; 370-377.
[38] Jurie, F., Triggs, B. Creating efficient codebooks for visual recognition [A]. In Proc. 10th IEEE Int'l Conf. Computer Vision[C], 2005; 604-610.
[39] Maree, R., Geurts, P., Piater, J., etc. Random subwindows for robust image classification [A]. In Proc. IEEE Conf. Computer Vision and Pattern Recognition[C], 2005; 34-40.
[40] Moosmann, F., Larlus, D., Jurie, F. Learning Saliency Maps for Object Categorization [A]. In Proc. European Conf. Computer Vision Workshop Representation and Use of Prior Knowlodge[C], 2006.
[41] Moosmann, F., Larlus, D., Jurie, F. Fast discriminative visual codebooks using randomized clustering forests [A]. In In Advances in Neural Information Processing Systems(Nips)[C], 2006; 985-992.
[42] Perronnin, F. Universal and adapted vocabularies for Generic Visual Categorization [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, Jul, 2008, 30 (7): 1243-1256.
[43] Yang, Y., Newsam, S. Bag-Of-Visual-Words and Spatial Extensions for Land-Use Classification [A]. In ACM SIGSPATIAL International Conference on Advances in Geographic InformationSystems[C], 2010.
[44] Jiang, Y. G., Ngo, C. W. Visual word proximity and linguistics for semantic video indexing and near-duplicate retrieval [J]. Computer Vision and Image Understanding 2009, 113 (3): 405-414.
[45] van Gemert, J. C., Veenman, C. J., Smeulders, A. W. M., etc. Visual Word Ambiguity [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, Jul, 2010, 32 (7): 1271-1283.
[46] Arthur Dempster, Nan Laird, and Donald Rubin. Maximum likelihood from incomplete data via the EM algorithm[J]. Journal of the Royal Statistical Society, Series B, 1977,39(1):1–38,
[47] Winn J M. Variational message passing and it's application[D]. Cambridge: Uinversity of Cambridge, 2003.
[48] Im J, Jensen J R. A change detection model based on neighborhood correlation image analysis and decision tree classification[J]. Remote Sensing of Enviornment, 2005, 99(3): 326-340.
[49] Xu Sheng, Fang Tao, Li Deren, Wang Shiwei. Object classification of aerial image with Bag-of-Visual-Words[J]. IEEE Geoscience and Remote Sensing Letters, 2010, 7(2): 366-370.