..............................
..............................
..............................
Zernike Moments and SVM for Shape Classification in Very High Resolution
In this paper, a Zernike moments"based descriptor i s used as a measure of shape information for the de tection of
buildings from Very High Spatial Resolution (VHSR) satellite images. The proposed approach comprises three steps. First, the
image is segmented into homogeneous objects based o n the spectral and spatial information. MeanShift segmentation method
is used for this end. Second, a Zernike feature vec tor is computed for each segment. Finally, a Suppor t Vector Machines
(SVM)"based classification using the feature vector s as inputs is performed. Experimental results and comparison with
Environment for Visualizing Images (ENVI)
commercial package confirm the effectiveness of the proposed approach.
[1] Abdel0Qader H., Ramli A., and Al0Haddad S., Fingerprint Recognition using Zernike Moments, International Arab Journal of Information Technology , vol. 4, no. 4, pp. 3720 376, 2007.
[2] Bouziani M., Goita K., and He D., Rule0based Classification of a Very High Resolution Image in an Urban Environment using Multispectral Segmentation Guided by Cartographic Data, IEEE Transaction on Geoscience and Remote Sensing , vol. 48, no. 8, pp. 319803211, 2010.
[3] Bruzzone L. and Carlin L., A Multilevel Context0Based System for Classification of Very High Spatial Resolution Images, IEEE Transaction on Geoscience and Remote Sensing , vol. 44, no. 9, pp. 258702600, 2006.
[4] Camps0Valls G., Gomez0Chova L., Calpe0 Maravia J., Martin0Guerrero J., Soria0Olivas E., Alonso0Chorda L., and Moreno J., Robust Support Vector Method for Hyperspectral Data Classification and Knowledge Discovery, IEEE Transaction on Geoscience and Remote Sensing , vol. 42, no. 7, pp. 153001542, 2004.
[5] Celebi M. and Aslandogan Y., A Comparative Study of Three Moment0based Shape Descriptors, in Proceedings of IEEE International Conference on Information Technology: Coding and Computing , Las Vegas, vol. 1, pp. 7880793, 2005.
[6] Chabrier S., Emile B., Laurent H., Rosenberger C., and March P., Unsupervised Evaluation of Image Segmentation: Application to Multispectral Images, in Proceedings of the 17 th International Conference Pattern Recognition , vol. 1, pp. 5760 579, 2004.
[7] Cheng Y., Mean Shift, Mode Seeking, and Clustering, IEEE Transaction on Pattern Analysis and Machine Intelligence , vol. 17, no. 8, pp. 7900799, 1995.
[8] Chong C., Raveendran P., and Mukundan R., Translation Invariants of Zernike Moments, Pattern Recognition , vol. 36, no. 8, pp. 17650 1773, 2003.
[9] Chong C., Raveendran P., and Mukundan R., A Comparative Analysis of Algorithms for Fast Computation of Zernike Moments, Pattern Recognition , vol. 36, no. 3, pp. 7310742, 2003.
[10] Comaniciu D. and Meer P., Mean Shift: A Robust Approach Towards Feature Space Analysis, IEEE Transaction on Pattern Analysis and Machine Intelligence , vol. 24, no. 5, pp. 6030619, 2002.
[11] Cristianini N. and Shaew0Taylor J., An Introduction to Support Vector Machines and other Kernel Based Learning Methods , Cambridge University Press, UK, 2000.
[12] Fukunaga K. and Hostetler L., The Estimation of the Gradient of a Density Function, with Applications in Pattern Recognition, IEEE Transaction on Information Theory , vol. 21, no. 1, pp. 32040, 1975.
[13] Gope C., Kehtarnavaz N., and Hillman G., Zernike Moment Invariants based Photo0 Identification using Fisher Discriminant Model, in Proceedings of the 26 th Annual International Conference of the IEEE Engineering in Medicine and Biology Society , San Francisco, vol. 1, pp. 145501458, 2004.
[14] Herold M., Gardner M., and Roberts D., Spectral Resolution Requirement for Mapping Urban Areas, IEEE Transaction on Geoscience and Remote Sensing , vol. 41, no. 9, pp. 19070 1919, 2003.
[15] Hofmann P., Detecting Urban Features from IKONOS Data using an Object0Oriented Approach, in Proceedings of the Remote Sensing and Photogrammetry Society , pp. 79091, 2003.
[16] Hu M., Visual Pattern Recognition by Moment Invariants, IRE Transaction on Information Theory , vol. 8, no. 2, pp. 1790187, 1962.
[17] Inglada J. and Michel J., Qualitative Spatial Reasoning for High0Resolution Remote Sensing Image Analysis, IEEE Transaction on Geoscience and Remote Sensing , vol. 47, no. 2, pp. 5990612, 2009.
[18] Kim H. and Lee H., Invariant Image Watermark using Zernike Moments, IEEE Transaction on Circuits and Systems for Video Technology , vol. 13, no. 8, pp. 7660775, 2003.
[19] Li S., Lee M., and Pun C., Complex Zernike Moments Features for Shape0Based Image Retrieval, IEEE Transaction on Systems, Man, and Cybernetics"Part A: Systems and Humans , vol. 39, no. 1, pp. 2270237, 2009.
[20] Loncaric S., A Survey of Shape Analysis Techniques, Pattern Recognition , vol. 31, no. 8, pp. 98301001, 1998.
[21] Nixon M. and Aguado A., Feature Extraction and Image Processing , Academic Press, UK, 2008.
[22] Persoon E. and Fu K., Shape Discrimination Zernike Moments and SVM for Shape Classification in Very High Resolution Satellite Images using Fourier Descriptors, IEEE Trans Systems, Man, and Cybernetics, vol 1700179, 1977.
[23] Shackelford A. and Davis C. , A Fuzzy Pixel0Based and Object0Based Approach for Classification of High Resolution Multispectral Data Over Urban Areas Transaction on Geoscience and Remote Sensing vol. 41, no. 10, pp. 235402363, 2003
[24] Teague M., Image Analysis v Theory of Moments, Journal of the Society of America, vol. 70, no. 1980.
[25] Teh C. and Chin R., On Image Methods of Moments, IEEE Transaction on Pattern Analysis and Machine Intelligence 10, no. 4, pp. 4960513, 1988.
[26] Vapnik V., Statistical Learning Theory New York, 1998.
[27] Zhang D. and Lu G., Review of Representation and Description Techniques Pattern Recognition , vol. 37, no 2004.
[28] Zhang D. and Lu G., Content Retrieval using Different Shape Descriptors Comparative Study, in Proceedings of International Conference on Multimedia and Expo , Japan, pp. 3170320, 2001. Habib Mahi Engineering degree science from the University of Mohamed Boudiaf 1994 and MS degree in Techniques Spatiales et Applications Centre of Space Techniques Algeria, in 2004. During 2004, he is a permanent researcher at the Earth Observation Division of CTS His research interests include: image processing, feature extraction, classification development for remotely sensed data. Hadria Isabaten Engineering degree i Engineering from the University of Mohamed Boudiaf Docteur Ing nieur industrial computer science automation from the University of Lille 1 in France, in 1987 and the computer science from the University of Mohamed Boudiaf in 2005. Currently she is a Faculty of Computer Science in the Mohamed Boudiaf. Her research interests cover image processing and the application of pattern recogniti on to remotely sensed data. Zernike Moments and SVM for Shape Classification in Very High Resolution Satellite Images IEEE Transaction on vol. 7, no. 3, pp. , A Combined Based Approach High Resolution Multispectral Data Over Urban Areas, IEEE Geoscience and Remote Sensing, 2003. via the General ournal of the Optical . 8, pp. 9200930, Image Analysis by the IEEE Transaction on Analysis and Machine Intelligence, vol. Learning Theory, Wiley, Zhang D. and Lu G., Review of Shape Description Techniques, no. 1, pp. 1019, and Lu G., Content0Based Shape Different Shape Descriptors: A in Proceedings of IEEE International Conference on Multimedia and received his degree in computer the University of in Algeria, in degree in Techniques Spatiales et Applications from the Centre of Space Techniques in he is a permanent researcher at the Earth Observation Division of CTS. research interests include: image processing, image feature extraction, classification algorithms . Isabaten received her degree in Electrical Engineering from the University of Mohamed Boudiaf in 1981, the Docteur Ing nieur degree in industrial computer science and from the University of and the phD degree in from the University of Mohamed a professor in the in the University of Her research interests cover image processing and the application of pattern recogniti on to Chahira Serief Engineer in Electrical University of Constantine in Algeria, respectively been working as a permanent researcher at the Earth Observation Division of the Centre of Space Techniques in Algeria. Her principal research interests are in the fields of satellite and medical image processing, image feature extraction; image registration, image segmentation and contour detection. 51 Chahira Serief received her Engineering, Ms, and PhD degrees in Electrical Engineering from the University of Constantine in , in 1995, 2000, and 2009 respectively. Since 2002 she has been working as a permanent researcher at the Earth Observation Division of the Centre of Space Techniques in Algeria. Her principal ch interests are in the fields of satellite and medical image processing, image feature extraction; image registration, image segmentation and contour