A LEARNING VECTOR QUANTIZATION FOR RECOGNITION OF INVARIANT SPATIAL DETECTORS OF PERCEPTUAL PATTERNS CONSTITUENTS

Hanan Hassan Ali Adian; Abd. Rahman Ramli; Bachok Taib; Adznan Jantan

Authors

Hanan Hassan Ali Adian
Abd. Rahman Ramli Head of Intelligent Systems and Robotics Laboratory (ISRL), Institute of Advanced Technology (ITMA), University Putra Malaysia
Bachok Taib Department of Mathematics, Faculty of Science, University Pendidikan Sultan Idris
Adznan Jantan Department of Computer and Communication Systems Engineering, Faculty of Engineering, University Putra Malaysia

Keywords:

LVQ, invariant recognition, spatial characteristics, satellite imagery

Abstract

This paper investigates the use of a learning vector quantization (LVQ) neural network trained by invariants and spatial detectors for pattern recognition. Satellite imagery patterns are sensitive to translation, rotation, and scale variability. This motivates the construction of such detectors to constitute perceptual instances for an LVQ network. Influence of certain factors such as the hidden layer neurons, and the learning parameter are investigated.

References

Gualtieri, J.A. (1988). Goddard researchers simulate neural networks using parallel processing. NASA Information Systems Newsletter, 15:12-15.

Hara, Y., Atkins, R.G., Yueh, S.H., Shin, R.T., and Kong, J.A. (1994). Application of neural networks to radar image classification. IEEE Transactions on Geoscience and Remote Sensing, 32(1):100-109.

Kamgar-Parsi, B., Gualtieri, J.A., and Devaney, J.E. (1990). Clustering with neural networks. Biological Cyberbetics, 63:201-208.

Kanellopoulos, I. (1997). Use of neural networks for improving satellite image processing techniques for land cover/ land use classification. Italy: European Commission, Joint Research Center. Available from: http://europa.eu.int. Accessed June, 2001.

Kanellopoulos, I., Wilkinson, G.G., and Chiuderi, A. (1994). Land cover mapping using combined Landsat TM imagery and textural features from ERS-1 Synthetic Aperture Radar Imagery. In: Image and Signal Processing for Remote Sensing. Desachy, J., ed., p. 332-341.

Knok, R., Hara, Y., Atkins, R.G., Yueh, S.H., Shin, R.T., and Kong, J.A. (1991). Application of neural networks to sea ice classification using polarimetric SAR image. Proceedings of the International Geoscience and Remote Sensing Symposium (IGARSS’91); June 1991; Espoo, Finland. IEEE Press, 1:85-88.

Looney, C.G. (1997). Pattern recognition using neural networks. Theory and Algorithme for Engineer and Scientists’ Oxford.

Luttrell, S.P. (1998) Image compression using a neural network. Proceedings of the International Geoscience and Remote Sensing Symposium, (IGASS’88); September,1998; Edinburgh, Scotland. ESA Publications Division, p. 1,231-1,238.

Milan, S., Hlavac, V., and Boyle, R. (1999). Image processing, analysis, and machine vision. ITP.

Webber, C. (1994). Self-organization of transformation invariant detectors for constituents of perceptual pattern. Network: Computation in Neural Syst., 5:471-496.

Wood, J. (1996) Invariant pattern recognition: A review’ pattern recognition. Elsevier Science, 29(1):1-17.