Abstract
This paper analyzes the application of the High Order Artificial Neural Networks (HOANN) to recognition of patterns invariant to projectivity transformation. This invariance is necessary in those applications in which either the objects or the own vision system can be displaced or turned in the environment. The invariant relation incorporated to the structure of our neural system is the cross-ratio of sines of angles of five coplanar points [Barrett, 1991]. This relation limits the application of these networks to the recognition of flat objects or threedimensional objects which are characterized by some plane of symmetry. As application we introduce a high order neural architecture applied to the projectivity invariant discrimination of projective transformed patterns of different modern aircraft.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
References
Barret, E. B. and Payton, P. M., (1991), “General Methods for Determining Projective Invariants in Imagery”, CVGIP: Image Understanding, Vol. 53, No. 1, pp. 46–65.
Giles,C.L. y Maxwell, T., (1987), “Learning, invariance, and generalization in high-order neural networks”, Applied Optics, Vol. 26, No. 23, pp. 4972–4978.
Gros, P.,(1993), “How to Use the Cross Ratio to Compute Projective Invariants from Two Images”, in Applications of Invariance in Computer Vision, J.L.Mundy, A. Zisserman and D. Forsyth (Eds), LNCS 825, Spriger-Verlag, pp. 107–126.
Joya, G., y Sandoval, F.,(1993), “Projectivity invariant pattern Recognition with High-Order Neural Networks”, in New Trends in Neural Computation, J. Mira, J. Cabestany y A. Prieto (Eds), LNCS 686 Springer-Verlag, pp.513–518.
Maybank, S.J.,(1993), “Classification Based on the Cross Ratio”, in Applications of Invariance in Computer Vision, J.L.Mundy, A. Zisserman and D. Forsyth (Eds), LNCS 825, Spriger-Verlag, pp. 453–472.
Perantonis, S.J. y Lisboa, P.J.G., (1992), “Translation, Rotation, and Scale Invariant Pattern Recognition by High-Order Neural Networks and Moment Classifiers”, IEEE Transactions on Neural Networks, Vol 3 No 2 pp 241–251.
Prado, J. M., (1986), “Guia Ilustrada de Cazas Modernos”, Ediciones Orbis S.A.
Reid, M. B., Spirkovska, L. y Ochoa, E., (1989), “Simultaneous position, scale, and rotation invariant pattern classification using third-order neural networks”, Neural Networks, Vol. 1, No. 3, pp. 154–159.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Joya, G., Sandoval, F. (1997). Neural networks based projectivity invariant recognition of flat patterns. In: Mira, J., Moreno-Díaz, R., Cabestany, J. (eds) Biological and Artificial Computation: From Neuroscience to Technology. IWANN 1997. Lecture Notes in Computer Science, vol 1240. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0032563
Download citation
DOI: https://doi.org/10.1007/BFb0032563
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63047-0
Online ISBN: 978-3-540-69074-0
eBook Packages: Springer Book Archive