This Article 
 Bibliographic References 
 Add to: 
Recognition by Linear Combinations of Models
October 1991 (vol. 13 no. 10)
pp. 992-1006

An approach to visual object recognition in which a 3D object is represented by the linear combination of 2D images of the object is proposed. It is shown that for objects with sharp edges as well as with smooth bounding contours, the set of possible images of a given object is embedded in a linear space spanned by a small number of views. For objects with sharp edges, the linear combination representation is exact. For objects with smooth boundaries, it is an approximation that often holds over a wide range of viewing angles. Rigid transformations (with or without scaling) can be distinguished from more general linear transformations of the object by testing certain constraints placed on the coefficients of the linear combinations. Three alternative methods of determining the transformation that matches a model to a given image are proposed.

[1] Y. S. Abu-Mostafa and D. Pslatis, "Optical neural computing,"Sci. Amer., vol. 256, pp. 66-73, 1987.
[2] R. Bajcsy and F. Solina, "Three dimensional object representation revisited," inProc. 1st ICCV Conf.(London), 1987, pp. 231-240.
[3] R. Basri and S. Ullman, "The alignment of objects with smooth surfaces," inProc. 2nd ICCV Conf.1988, pp. 482-488.
[4] I. Biederman, "Human image understanding: Recent research and a theory,"Comput. Vision Graphics Image Processing, vol. 32, pp. 29-73, 1985.
[5] C. H. Chien and J. K. Aggarwal, "Shape recognition from single silhouette," inProc. ICCV Conf.(London), 1987, pp. 481-490.
[6] R. W. Brockett, "Least squares matching problems," inLinear Algebra Appl., pp. 1-17, 1989.
[7] O. D. Faugeras and M. Hebert, "The representation, recognition, and locating of 3-D objects,"Int. J. Robotics Res., vol. 5, no. 3, Fall 1986, pp. 27-52.
[8] M. A. Fischler and R. C. Bolles, "Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography,"Commun. ACM, vol. 24, no. 6, pp. 381-395, 1981.
[9] W. E. L. Grimson, "The combinatorics of heuristic search termination for object recognition in cluttered environments," O. Faugeras (Ed.),Proceedings ECCV 1990(Berlin).
[10] W. E. L. Grimson and T. Lozano-Perez, "Model-based recognition and localization from sparse data,"Int. J. Robotics Res., vol. 3, pp. 3-35, 1984.
[11] T. S. Huang and C. H. Lee, "Motion and structure from orthographic projections,"IEEE Trans. Patt. Anal. Machine Intell., vol. 12, no. 5, pp. 536-540, 1989.
[12] D. P. Huttenlocher and S. Ullman, "Object recognition using alignment," inProc. ICCV Conf.(London), 1987, pp. 102-111.
[13] J. J. Koenderink and A.J. Van Doorn, "The internal representation of solid shape with respect to vision,"Biol. Cybernetics, vol. 32, pp. 211-216, 1989; in G.E. Hinton and J.A. Anderson,Parallel Models of Associative Memory.Hillsdale, NJ: Lawrence Erlbaum, pp. 105-143.
[14] Y. Lamdan, J.T. Schwartz, and H. Wolfson, "On recognition of 3-D objects from 2-D images,"Courant Inst. Math. Sci., Robotics Tech. Rep. 122, 1987.
[15] C.H. Longuet-Higgins, "A computer algorithm for reconstructing a scene from two projections," Nature, vol. 293, pp. 133-135, 1981.
[16] D. Lowe, Perceptual Organization And Visual Recognition. Boston: Kluwer, 1985.
[17] D. Marr, "Analysis of occluding contour,"Phil. Trans. R. Soc. Lond. B, vol. 275, pp. 483-524, 1977.
[18] D. Marr and S. Ullman, "Directional selectivity and its use in early visual processing," inProc. R. Soc. Lond. B, vol. 211, pp. 151-180, 1981.
[19] D. Shoham and U. Ullman, "Aligning a model to an image using minimal information," inProc. 2nd ICCV Conf., 1988, pp. 259-263.
[20] D. Thompson and J. Mundy, "Three dimensional model matching from an unconstrained viewpoint,"Proc. Int. Conf. Robotics Automation, 1987, pp. 208-220.
[21] S. Ullman,The Interpretation of Visual Motion. Cambridge, MA: MIT Press, 1979.
[22] S. Ullman, "Recent computational studies in the interpretation of structure from motion," in A. Rosenfeld and J. Beck (Eds.),Human and Machine Vision. New York: Academic, 1983.
[23] S. Ullman, "Aligning pictorial descriptions: An approach to object recognition:Cognition, vol. 32, no. 3, pp. 193-254, 1989; A.I. Memo 931, Artificial Intell. Lab., Mass. Inst. Technol., 1986.
[24] H. Freeman and Chakravarty, "The use of characteristic views in the recognition of three-dimensional objects," in E. Gelsema and L. Kanal (Eds.),Pattern Recognition in Practice. Amsterdam: North-Holland, 1980.
[25] A. L. Yuille, D. S. Cohen, and P. W. Hllinan, "Feature extraction from faces using deformable templates," inProc. Comput. Vision Patt. Recogn., (San Diego), 1988, pp. 104-109.
[26] D. Zipser and R.A. Andersen, "A back-propagation programmed network that simulates response properties of a subset of posterior parietal neurons,"Nature, vol. 331, pp. 679-684, 1988.

Index Terms:
model combination; image combination; rigid transformations; linear combinations; visual object recognition; sharp edges; smooth bounding contours; computerised pattern recognition; computerised picture processing
S. Ullman, R. Basri, "Recognition by Linear Combinations of Models," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 13, no. 10, pp. 992-1006, Oct. 1991, doi:10.1109/34.99234
Usage of this product signifies your acceptance of the Terms of Use.