SAR IMAGE CLASSIFICATION USING SUPERVISED NEURAL CLASSIFIERS

An investigation about four supervised neural classifiers based on the Minkovski-r error and the modified Fisher criterion is evaluated to classify a double textured SAR amplitude image. Regions around preclassified pixels are presented to train the neural network that learns a sub-optimal set of masks via backpropagation algorithm. Classification performance is evaluated using kappa statistics. The neural classifiers showed almost the same performance for different window mask sizes and training samples. However, the Minkovski-r=1.1 error showed a slightly better performance than the others. Best results are obtained when the neural classified image is followed by an erosion process via Median filter. The results outperformed the classification performance of two statistical classifiers: the Minimum Bayes error and the Kullback-Liebler distance.

[1]  Hosni Ghedira,et al.  Application of neural networks for wetland classification in RADARSAT SAR imagery , 2000, IGARSS 2000. IEEE 2000 International Geoscience and Remote Sensing Symposium. Taking the Pulse of the Planet: The Role of Remote Sensing in Managing the Environment. Proceedings (Cat. No.00CH37120).

[2]  S. Quegan,et al.  Understanding Synthetic Aperture Radar Images , 1998 .

[3]  Jon Atli Benediktsson,et al.  Neural Network Approaches Versus Statistical Methods in Classification of Multisource Remote Sensing Data , 1989, 12th Canadian Symposium on Remote Sensing Geoscience and Remote Sensing Symposium,.

[4]  Stephen E. Fienberg,et al.  Discrete Multivariate Analysis: Theory and Practice , 1976 .

[5]  Kun-Shan Chen,et al.  Classification of multifrequency polarimetric SAR imagery using a dynamic learning neural network , 1996, IEEE Trans. Geosci. Remote. Sens..

[6]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[7]  David Fernandes,et al.  SAR image classification using a neural classifier based on Fisher criterion , 2002, VII Brazilian Symposium on Neural Networks, 2002. SBRN 2002. Proceedings..

[8]  Kun-Shan Chen,et al.  A fuzzy neural network to SAR image classification , 1998, IEEE Trans. Geosci. Remote. Sens..

[9]  Simon Yueh,et al.  Application of neural networks to radar image classification , 1994, IEEE Trans. Geosci. Remote. Sens..

[10]  Sigeru Omatu,et al.  A polarimetric SAR data classification method using neural networks , 1998, IGARSS '98. Sensing and Managing the Environment. 1998 IEEE International Geoscience and Remote Sensing. Symposium Proceedings. (Cat. No.98CH36174).

[11]  Horst Bischof,et al.  Multispectral classification of Landsat-images using neural networks , 1992, IEEE Trans. Geosci. Remote. Sens..

[12]  D. Solimini,et al.  Monitoring urban areas by using ERS-SAR data and neural networks algorithms , 1999, IEEE 1999 International Geoscience and Remote Sensing Symposium. IGARSS'99 (Cat. No.99CH36293).