Abstract
A novel architecture for real time pick up of 3D motion and 3D layout information of the visible surface of the rigid world, from the flow of the optic array, is presented.
The flow of the optic array contains the information of the motion of the physical world due to its “specificity” of the coherent changes. Such “specificity” is a one to one correspondence of 3D motions of a visible surface and its induced transformations of images.
The neural network architecture we propose is of the Pitts and McCulloch type. However, based on the second canonical coordinate system of the Lie transformation group, to facilitate image transformations induced by rigid motion of the visible surfaces in three dimensional space, one only needs 12 N intermediate nodes, where N is the number of neurons in the image plane. Thus the intractable complexity of Pitts and McCulloch’s neural network is overcome.
Research supported by grant DCR-8504011 from the Intelligent Systems Program, U.S. National Science Foundation.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Belinfante, J.G.; Kolman, B. and Smith, H.A. [66] “An Introduction to Lie Groups and Lie Algebras, With Applications”, in: SIAM Review Vol. 8, No. 1, January, 1966 11–46.
Braunstein, M.L. [76], Depth Perception Through Motion New York: Academic Press 1976.
Gibson, J.J. [66], The Senses Considered as Perceptual Systems Boston: Houghton Mifflin Company 1966.
Gibson, J.J. [79], The Ecological Approach to Visual Perception Boston: Houghton Mifflin Company 1979.
Hoffman, W.C. [64], “Pattern Recognition and The Method of Isoclines,” Boeing Scientific Report, Boeing Scientific Research Laboratories, Seattle, 1964.
Hoffman, W.C. [68], “The Neuron As A Lie Group Germ and A Lie Product”, in: Quarterly of Applied Mathematics Vol. XXV, No. 4, January, 1968 433–440.
Horn, B.K.P. and Schunck, B.G. [80], “Determining Optical Flow,” in: Artificial Intelligence, 11 (1980) 185–203.
Longuet-Higgins, H.C. and Prazdny, K. [80], “The Interpretation of A Moving Retinal Image”, in: Proc. Roy. Soc. B 208 (1980) 385–397
Longuet-Higgins, H.C. [84], “The Visual Ambiguity of A Moving Plane”, in: Proc. Roy. Soc. B 223 (1984)
Matsushima, Yozo [72], Differentiable Manifolds Marcel Dekker, Inc. New York 1972.
Pitts, W. and McCulloch, W.S. [47] “How We Know Universal, The Perception of Auditory and Visual Forms” Bulletin of Mathematical Biophysics Vol. 9 (1947) pp. 127–147.
Pontrjagin, L. [46], Topological Groups, Princeton University Press, Princeton 1946, 84.
Tsai, R.Y. and Huang, T.S. [84], “Uniqueness and Estimation of Three Dimensional Motion Parameters of Rigid Objects with Curved Surfaces”, in: IEEE Transactions on PAMI 6 1984 13–26.
Tsao, T. and Kanal, L. [87] “A Lie Group Approach to Visual Perception,” TR. 1852, Computer Science Department, Univ. of Maryland. May (1987).
Ullman S. [79], The Interpretation of Visual Motion (MIT Press, 1979).
Ullman S. [84], “Maximizing The Rigidity: The Incremental Recovery of 3-D Structure from Rigid and Rubbery Motion”, in: Perception, 13, 1984 255–274.
Wallach, H. and O’Connell D.N. [53], “The Kinetic Depth Effect”, in: Journal of Experimental Psychology 45 1953 205–217.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1988 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Tsao, T., Kanal, L. (1988). A Novel Architecture for Real Time Pick Up of 3D Motion and 3D Layout Information from The Flow of The Optic Array. In: Jain, A.K. (eds) Real-Time Object Measurement and Classification. NATO ASI Series, vol 42. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-83325-0_17
Download citation
DOI: https://doi.org/10.1007/978-3-642-83325-0_17
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-83327-4
Online ISBN: 978-3-642-83325-0
eBook Packages: Springer Book Archive