Impervious surface extraction in imbalanced datasets: integrating partial results and multi-temporal information in an iterative one-class classifier

ABSTRACT Accurate urban land use/cover monitoring is an essential step towards a sustainable future. As a key part of the classification process, the characteristics of reference data can significantly affect classification accuracy and quality of produced maps. However, ideal reference data is not always readily available; users frequently have difficulty generating sufficient reference data for some classes given time, cost, data availability, expertise level, or other limitations. This study aims at dealing with this lack of sufficiently balanced reference data by presenting a modified hybrid one-class support vector data description (SVDD) model. The underlying hypothesis is that the lack of balanced reference data can be overcome through integration of partially extracted results and multi-temporal spectral information. The partially extracted results, defined as highly accurate classified pixels identified in previous algorithmic iterations, allow a gradual increase of the available training data. Furthermore, the method incorporates a voting system that integrates multi-temporal images using the SVDD algorithm. We applied this hybrid method to binary impervious classification of multi-temporal Landsat Thematic Mapper imagery from Central New York with imbalanced reference data. The proposed hybrid one-class SVDD model achieved a 5–6% improvement in overall accuracy and 0.05–0.09 in kappa than the typical one-class SVDD benchmark. While the method was tested on a single site (albeit with an unusually high reference dataset size of >870,000 pixels) we feel confident to suggest implementation of our methodology in other sites over the traditional method. This is because our approach automatically reverts to the traditional method when voting is inconsistent or there is a limited number of highly accurately classified pixels to assist future iterations. Future work could explore the quantity and temporal specificity (e.g. benefits of specific months) of the multi-temporal image selection and/or test other one-class classifiers.

[1]  Xiaojun Yang,et al.  Estimating landscape imperviousness index from satellite imagery , 2006, IEEE Geosci. Remote. Sens. Lett..

[2]  Johannes R. Sveinsson,et al.  Classifying Remote Sensing Data with Support Vector Machines and Imbalanced Training Data , 2009, MCS.

[3]  Randel L. Dymond,et al.  Evaluation of Impervious Surface Estimates in a Rapidly Urbanizing Watershed , 2004, Photogrammetric Engineering & Remote Sensing.

[4]  Andrew Chan,et al.  An SVDD-Based Algorithm for Target Detection in Hyperspectral Imagery , 2011, IEEE Geoscience and Remote Sensing Letters.

[5]  V. Vapnik Estimation of Dependences Based on Empirical Data , 2006 .

[6]  Dengsheng Lu,et al.  Detection of impervious surface change with multitemporal Landsat images in an urban-rural frontier. , 2011, ISPRS journal of photogrammetry and remote sensing : official publication of the International Society for Photogrammetry and Remote Sensing.

[7]  Giorgos Mountrakis,et al.  Converting local spectral and spatial information from a priori classifiers into contextual knowledge for impervious surface classification , 2011 .

[8]  G. Mountrakis,et al.  Developing collaborative classifiers using an expert-based model. , 2009 .

[9]  Giorgos Mountrakis,et al.  Integration of urban growth modelling products with image-based urban change analysis , 2013 .

[10]  Jieping Ye,et al.  A Small Sphere and Large Margin Approach for Novelty Detection Using Training Data with Outliers , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[12]  Maggi Kelly,et al.  Support vector machines for predicting distribution of Sudden Oak Death in California , 2005 .

[13]  K. Beurs,et al.  Remote sensing of crop residue cover using multi-temporal Landsat imagery , 2012 .

[14]  Heesung Kwon,et al.  Sparse Kernel-Based Hyperspectral Anomaly Detection , 2012, IEEE Geoscience and Remote Sensing Letters.

[15]  A. Elmore,et al.  Synergistic use of Landsat Multispectral Scanner with GIRAS land-cover data to retrieve impervious surface area for the Potomac River Basin in 1975 , 2010 .

[16]  D. Civco,et al.  SUBPIXEL IMPERVIOUS SURFACE MAPPING , 2001 .

[17]  Luis Gómez-Chova,et al.  Combination of one-class remote sensing image classifiers , 2007, 2007 IEEE International Geoscience and Remote Sensing Symposium.

[18]  Haiqing Xu,et al.  Urban building damage detection from very high resolution imagery by One-Class SVM and shadow information , 2011, 2011 IEEE International Geoscience and Remote Sensing Symposium.

[19]  Michael Brady,et al.  Novelty detection for the identification of masses in mammograms , 1995 .

[20]  Xindong Wu,et al.  Support vector machines based on K-means clustering for real-time business intelligence systems , 2005, Int. J. Bus. Intell. Data Min..

[21]  Robert P. W. Duin,et al.  Support vector domain description , 1999, Pattern Recognit. Lett..

[22]  John A. Richards,et al.  Remote Sensing Digital Image Analysis , 1986 .

[23]  Kun Jia,et al.  Multi-temporal remote sensing data applied in automatic land cover update using iterative training sample selection and Markov Random Field model , 2015 .

[24]  J. R. Jensen,et al.  Effectiveness of Subpixel Analysis in Detecting and Quantifying Urban Imperviousness from Landsat Thematic Mapper Imagery , 1999 .

[25]  Shobha V. Iyer,et al.  Urban landuse monitoring using neural network classification , 2002, IEEE International Geoscience and Remote Sensing Symposium.

[26]  Nello Cristianini,et al.  Kernel Methods for Pattern Analysis , 2003, ICTAI.

[27]  Robert M. Haralick,et al.  Textural Features for Image Classification , 1973, IEEE Trans. Syst. Man Cybern..

[28]  Giorgos Mountrakis,et al.  Enhancing and replacing spectral information with intermediate structural inputs: A case study on impervious surface detection , 2011 .

[29]  Qihao Weng,et al.  Remote sensing of impervious surfaces in the urban areas: Requirements, methods, and trends , 2012 .

[30]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[31]  M. Ridd Exploring a V-I-S (vegetation-impervious surface-soil) model for urban ecosystem analysis through remote sensing: comparative anatomy for cities , 1995 .

[32]  J. Paruelo,et al.  Land cover classification in the Argentine Pampas using multi-temporal Landsat TM data , 2003 .

[33]  Sangbum Lee,et al.  Subpixel analysis of Landsat ETM/sup +/ using self-organizing map (SOM) neural networks for urban land cover characterization , 2006, IEEE Transactions on Geoscience and Remote Sensing.

[34]  Annalisa Appice,et al.  Iterative Hyperspectral Image Classification Using Spectral–Spatial Relational Features , 2015, IEEE Transactions on Geoscience and Remote Sensing.

[35]  A. Atiya,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2005, IEEE Transactions on Neural Networks.

[36]  Sungzoon Cho,et al.  EUS SVMs: Ensemble of Under-Sampled SVMs for Data Imbalance Problems , 2006, ICONIP.

[37]  Jing Liu,et al.  The Application of SVDD in Gene Expression Data Clustering , 2008, 2008 2nd International Conference on Bioinformatics and Biomedical Engineering.

[38]  Yuxiao Hu,et al.  One-class classification for spontaneous facial expression analysis , 2006, 7th International Conference on Automatic Face and Gesture Recognition (FGR06).

[39]  Giorgos Mountrakis,et al.  A meta-analysis of remote sensing research on supervised pixel-based land-cover image classification processes: General guidelines for practitioners and future research , 2016 .

[40]  Ryutaro Tateishi,et al.  A hybrid pansharpening approach and multiscale object-based image analysis for mapping diseased pine and oak trees , 2013 .

[41]  Corina da Costa Freitas,et al.  Mapping impervious surfaces with the integrated use of Landsat Thematic Mapper and radar data: A case study in an urban–rural landscape in the Brazilian Amazon , 2011 .

[42]  Martin Herold,et al.  Mapping imperviousness using NDVI and linear spectral unmixing of ASTER data in the Cologne-Bonn region (Germany) , 2004, SPIE Remote Sensing.

[43]  Zne-Jung Lee,et al.  Parameter determination of support vector machine and feature selection using simulated annealing approach , 2008, Appl. Soft Comput..

[44]  Giorgos Mountrakis,et al.  International Journal of Remote Sensing , 2022 .

[45]  Nathalie Japkowicz,et al.  The class imbalance problem: A systematic study , 2002, Intell. Data Anal..

[46]  Wenkai Li,et al.  A Positive and Unlabeled Learning Algorithm for One-Class Classification of Remote-Sensing Data , 2011, IEEE Transactions on Geoscience and Remote Sensing.

[47]  D. Lu,et al.  Use of impervious surface in urban land-use classification , 2006 .

[48]  Curt H. Davis,et al.  A combined fuzzy pixel-based and object-based approach for classification of high-resolution multispectral data over urban areas , 2003, IEEE Trans. Geosci. Remote. Sens..

[49]  Chih-Jen Lin,et al.  A Practical Guide to Support Vector Classication , 2008 .

[50]  Mohamed Cheriet,et al.  Iterative Classifiers Combination Model for Change Detection in Remote Sensing Imagery , 2016, IEEE Transactions on Geoscience and Remote Sensing.

[51]  Antonio F. Wolf,et al.  Using WorldView-2 Vis-NIR multispectral imagery to support land mapping and feature extraction using normalized difference index ratios , 2012, Defense + Commercial Sensing.

[52]  C.N. Manikopoulos,et al.  Windows NT one-class masquerade detection , 2004, Proceedings from the Fifth Annual IEEE SMC Information Assurance Workshop, 2004..

[53]  Giles M. Foody,et al.  Training set size requirements for the classification of a specific class , 2006 .

[54]  Ian Dear,et al.  Autonomously Determining the Parameters for SVDD with RBF Kernel from a One-Class Training Set , 2013 .

[55]  Marvin E. Bauer,et al.  ESTIMATION, MAPPING AND CHANGE ANALYIS OF IMPERVIOUS SURFACE AREA BY LANDSAT REMOTE SENSING , 2005 .

[56]  Carsten Jürgens,et al.  Seasonal multitemporal land-cover classification and change detection analysis of Bochum, Germany, using multitemporal Landsat TM data , 2016 .

[57]  Chien-Shun Lo,et al.  Support vector machine for breast MR image classification , 2012, Comput. Math. Appl..

[58]  Francesca Bovolo,et al.  Semisupervised One-Class Support Vector Machines for Classification of Remote Sensing Data , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[59]  R. Madlener,et al.  Impacts of urbanization on urban structures and energy demand: What can we learn for urban energy planning and urbanization management? , 2011 .

[60]  T. Carlson,et al.  On the relation between NDVI, fractional vegetation cover, and leaf area index , 1997 .

[61]  D. Roberts,et al.  Sub-pixel mapping of urban land cover using multiple endmember spectral mixture analysis: Manaus, Brazil , 2007 .

[62]  Mike Cranea,et al.  ESTIMATION OF SUB-PIXEL IMPERVIOUS SURFACES USING LANDSAT AND ASTER IMAGERY FOR ASSESSING URBAN GROWTH , 2005 .

[63]  Giorgos Mountrakis,et al.  Integrating intermediate inputs from partially classified images within a hybrid classification framework: An impervious surface estimation example , 2010 .

[64]  Robert P. W. Duin,et al.  Uniform Object Generation for Optimizing One-class Classifiers , 2002, J. Mach. Learn. Res..

[65]  Francisco Herrera,et al.  Addressing the Classification with Imbalanced Data: Open Problems and New Challenges on Class Distribution , 2011, HAIS.

[66]  Limin Yang,et al.  An approach for mapping large-area impervious surfaces: synergistic use of Landsat-7 ETM+ and high spatial resolution imagery , 2003 .

[67]  David S. Doermann,et al.  Support Vector Data Description for image categorization from Internet images , 2008, 2008 19th International Conference on Pattern Recognition.

[68]  Amit Banerjee,et al.  A support vector method for anomaly detection in hyperspectral imagery , 2006, IEEE Transactions on Geoscience and Remote Sensing.

[69]  Jungho Im,et al.  ISPRS Journal of Photogrammetry and Remote Sensing , 2022 .

[70]  Chou-Yuan Lee,et al.  A novel algorithm applied to classify unbalanced data , 2012, Appl. Soft Comput..

[71]  Wu Zhaohui,et al.  Support vector domain description for speaker recognition , 2001, Neural Networks for Signal Processing XI: Proceedings of the 2001 IEEE Signal Processing Society Workshop (IEEE Cat. No.01TH8584).

[72]  B. Gao NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space , 1996 .

[73]  Malik Yousef,et al.  One-Class SVMs for Document Classification , 2002, J. Mach. Learn. Res..

[74]  Alan T. Murray,et al.  Estimating impervious surface distribution by spectral mixture analysis , 2003 .

[75]  Xue-wen Chen,et al.  Comparison of One-Class SVM and Two-Class SVM for Fold Recognition , 2006, ICONIP.

[76]  D. Civco,et al.  IMPERVIOUS SURFACE MAPPING FOR THE STATE OF CONNECTICUT 1 , 1997 .

[77]  Lei Zhang,et al.  Automatic land-cover update approach integrating iterative training sample selection and a Markov Random Field model , 2014 .

[78]  F. Canters,et al.  A comparison of two spectral mixture modelling approaches for impervious surface mapping in urban areas , 2009 .

[79]  Giorgos Mountrakis Next Generation Classifiers : Focusing on Integration Frameworks , 2008 .

[80]  Gunter Ritter,et al.  Outliers in statistical pattern recognition and an application to automatic chromosome classification , 1997, Pattern Recognit. Lett..

[81]  Andrew J. Smith,et al.  Subpixel Estimates of Impervious Surface Cover Using Landsat TM Imagery , 2002 .

[82]  Amit Banerjee,et al.  Fast Hyperspectral Anomaly Detection via SVDD , 2007, 2007 IEEE International Conference on Image Processing.

[83]  Margaret E. Gardner,et al.  Mapping Chaparral in the Santa Monica Mountains Using Multiple Endmember Spectral Mixture Models , 1998 .

[84]  Robert P. W. Duin,et al.  A Study On Combining Image Representations For Image Classification And Retrieval , 2004, Int. J. Pattern Recognit. Artif. Intell..

[85]  S. B. Serpico,et al.  Training of neural networks for classification of imbalanced remote-sensing data , 1997, IGARSS'97. 1997 IEEE International Geoscience and Remote Sensing Symposium Proceedings. Remote Sensing - A Scientific Vision for Sustainable Development.

[86]  Hanseok Ko,et al.  Face detection using support vector domain description in color images , 2004, 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[87]  Robert P. W. Duin,et al.  Support Vector Data Description , 2004, Machine Learning.

[88]  Marvin E. Bauer,et al.  MULTI-TEMPORAL LANDSAT IMAGE CLASSIFICATION AND CHANGE ANALYSIS OF LAND COVER IN THE TWIN CITIES (MINNESOTA) METROPOLITAN AREA , 2004 .

[89]  Nathaniel D. Herold,et al.  MAPPING IMPERVIOUS SURFACES AND FOREST CANOPY USING CLASSIFICATION AND REGRESSION TREE (CART) ANALYSIS , 2003 .

[90]  Saeid Homayouni,et al.  Improving the SVDD Approach to Hyperspectral Image Classification , 2012, IEEE Geoscience and Remote Sensing Letters.

[91]  Giles M. Foody,et al.  Sanchez-Hernandez, Carolina and Boyd, Doreen S. and Foody, Giles M. (2007) One-class classification for monitoring a specific land cover class: SVDD classification of fenland. IEEE Transactions on , 2016 .

[92]  Mayer Aladjem Recursive training of neural networks for classification , 2000, IEEE Trans. Neural Networks Learn. Syst..

[93]  T. Schueler The importance of imperviousness , 1995 .