Critical class sensitive active learning method for classification of remote sensing imagery

Remote sensing images provide essential data source for monitoring the land cover and land change on the Earth with a fast revisiting period. To fully utilize the remote sensing data, supervised classification methods are good choices to convert the data to land cover types due to their good abilities. One of the great challenges is to effectively collect training samples, especially for remote sensing images with an area scale or even global scale. One possible solution is using advanced machine learning techniques, e.g., active learning methods, to define training samples effectively and concisely. In this paper, we focus on critical class (i.e., the class which is hard to classify accurately) sensitive active learning methods for remote sensing image classification. The proposed algorithm is based on the widely-used support vector machines classifier. Experimental tests are performed on two public hyperspectral image data sets. Preliminary results show the effectiveness of the proposed algorithm.

[1]  David J. C. MacKay,et al.  Information-Based Objective Functions for Active Data Selection , 1992, Neural Computation.

[2]  William A. Gale,et al.  A sequential algorithm for training text classifiers , 1994, SIGIR '94.

[3]  Shlomo Argamon,et al.  Committee-Based Sampling For Training Probabilistic Classi(cid:12)ers , 1995 .

[4]  Nello Cristianini,et al.  Query Learning with Large Margin Classi ersColin , 2000 .

[5]  Greg Schohn,et al.  Less is More: Active Learning with Support Vector Machines , 2000, ICML.

[6]  J. Townshend,et al.  Global land cover classi(cid:142) cation at 1 km spatial resolution using a classi(cid:142) cation tree approach , 2004 .

[7]  Daphne Koller,et al.  Support Vector Machine Active Learning with Applications to Text Classification , 2000, J. Mach. Learn. Res..

[8]  Chih-Jen Lin,et al.  A comparison of methods for multiclass support vector machines , 2002, IEEE Trans. Neural Networks.

[9]  H. Sebastian Seung,et al.  Selective Sampling Using the Query by Committee Algorithm , 1997, Machine Learning.

[10]  G. Foody.  Thematic map comparison: Evaluating the statistical significance of differences in classification accuracy , 2004 .

[11]  Sankar K. Pal,et al.  Segmentation of multispectral remote sensing images using active support vector machines , 2004, Pattern Recognit. Lett..

[12]  Joydeep Ghosh,et al.  Investigation of the random forest framework for classification of hyperspectral data , 2005, IEEE Transactions on Geoscience and Remote Sensing.

[13]  Ishwar K. Sethi,et al.  Confidence-based active learning , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  E. Belle,et al.  Abstract , 2007 .

[15]  Joydeep Ghosh,et al.  An Active Learning Approach to Hyperspectral Data Classification , 2008, IEEE Transactions on Geoscience and Remote Sensing.

[16]  William J. Emery,et al.  Active Learning Methods for Remote Sensing Image Classification , 2009, IEEE Transactions on Geoscience and Remote Sensing.

[17]  P. Strobl,et al.  Pan-European Forest/Non-Forest Mapping with Landsat ETM+ and CORINE Land Cover 2000 Data , 2009 .

[18]  Melba M. Crawford,et al.  Critical class oriented active learning for hyperspectral image classification , 2011, 2011 IEEE International Geoscience and Remote Sensing Symposium.

[19]  Mikhail F. Kanevski,et al.  A Survey of Active Learning Algorithms for Supervised Remote Sensing Image Classification , 2011, IEEE Journal of Selected Topics in Signal Processing.

[20]  Lorenzo Bruzzone,et al.  Batch-Mode Active-Learning Methods for the Interactive Classification of Remote Sensing Images , 2011, IEEE Transactions on Geoscience and Remote Sensing.

[21]  Ping Tang,et al.  A Batch-Mode Active Learning Algorithm Using Region-Partitioning Diversity for SVM Classifier , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.