Support Vector Machine and Bhattacharrya kernel function for region based classification

Region based methods are indicated to classify image with strong heterogeneity, where only the spectral information is not enough. Different approaches have been proposed to perform this kind of classification. This study presents a new approach for region based classification that consists in use the Support Vector Machine (SVM) method with Bhattacharyya kernel function. A high resolution IKONOS image was classified. The classification results shows that SVM method using the Bhattacharyya kernel is better than Minimum Distance Classifier and conventional SVM.

[1]  Dong-Chul Park,et al.  Application of Bhattacharyya kernel-based Centroid Neural Network to the classification of audio signals , 2009, 2009 International Joint Conference on Neural Networks.

[2]  Fan Xia,et al.  Assessing object-based classification: advantages and limitations , 2009 .

[3]  Leandro Pardo,et al.  On the applications of divergence type measures in testing statistical hypotheses , 1994 .

[4]  Meritxell Bach Cuadra,et al.  Region-based satellite image classification: method and validation , 2005, IEEE International Conference on Image Processing 2005.

[5]  Sang Joon Kim,et al.  A Mathematical Theory of Communication , 2006 .

[6]  Renato J. Cintra,et al.  Hypothesis Testing in Speckled Data With Stochastic Distances , 2012, IEEE Transactions on Geoscience and Remote Sensing.

[7]  Volker Walter,et al.  Object-based classification of remote sensing data for change detection , 2004 .

[8]  Tony Jebara,et al.  A Kernel Between Sets of Vectors , 2003, ICML.

[9]  Bernhard Schölkopf,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2005, IEEE Transactions on Neural Networks.

[10]  Nuno Vasconcelos,et al.  A Kullback-Leibler Divergence Based Kernel for SVM Classification in Multimedia Applications , 2003, NIPS.

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

[12]  Corina da Costa Freitas,et al.  Land cover discrimination at Brazilian Amazon using region based classifier and stochastic distance , 2011, 2011 IEEE International Geoscience and Remote Sensing Symposium.

[13]  Sergios Theodoridis,et al.  Pattern Recognition, Fourth Edition , 2008 .