Object Recognition: Solution of the Simultaneous Pose and Correspondence Problem. Reconnaissance D’Objets Volumiques par Mise en Correspondance D’Indices Visuels

Object Recognition: Solution of the Simultaneous Pose and Correspondence Problem

Reconnaissance D' Objets Volumiques par Mise en Correspondance D' Indices Visuels

Frédéric Jurie

LASMEA – UMR 6602 du CNRS, Campus Universitaire des Cézeaux, 63177 Aubière Cedex

Page: 
321-344
|
Received: 
2 October 2000
|
Accepted: 
N/A
|
Published: 
31 December 2001
| Citation

OPEN ACCESS

Abstract: 

The use of hypothesis verification is recurrent in the model-based recognition literature. Verification consists in measuring how many model features transformed by a pose coincide with some image features. When data involved in the computation of the pose are noisy, the pose is inaccurate and difficult to verify, especially when the objects are partially occluded. To address this problem, the noise in image features is modeled by a Gaussian distribution. A probabilistic framework allows the evaluation of the probability of a matching, knowing that the pose belongs to a rectangular volume of the pose space. It involves quadratic programming, if the transformation is affine. This matching probability is used in an algorithm computing the best pose. It consists in a recursive multi resolution exploration of the pose space, discarding outliers in the match data while the search is progressing. Numerous experimental results are described. They consist of 2D and 3D recognition experiments using the proposed algorithm.

Résumé

Nous nous intéressons à la reconnaissance d’objets volumiques par mise en correspondance d’indices visuels. Nous supposons que les objets à reconnaître sont représentés à l’aide de modèles tridimensionnels, composés d’indices visuels. Reconnaître un objet signifie, dans ce cas, mettre en correspondance les indices du modèle de cet objet avec des indices extraits de l’image, de manière à ce que ces derniers puissent s’expliquer comme une transformation géométrique des indices du modèle. La recherche de la pose (valeur des paramètres de la transformation alignant le modèle sur l’image) et la recherche des correspondances sont ici traitées simultanément. Cela constitue l’originalité et la force de la méthode que nous proposons. Nous présentons de nombreux résultats expérimentaux illustrant l’utilisation de notre approche pour la reconnaissance d’objets.

Keywords: 

Model-based recognition; pose verificatio

Mots clés 

Reconnaissance d’objets, mise en correspondance

1. Introduction
2. Vers une Technique Efficace pour la Mise en Correspondance
3. Exploration Récursive de L’Espace des Poses
4. Probabilité qu’un Modèle d’Objet Soit dans l’Image, pour une Pose Donnée, et pour une Boîte Donnée.
5. Probabilité d’un Appariement Primitive Modèle – Primitive Image Connaissant une Pose p P(M|p)
6. Probabilité d’une Correspondance de Primitive Image à Primitive Modèle pour une Boîte Donnée (P(C|BOX))
7. Application à la Reconnaissance d’Objets par Utilisation de Collections de Vues 2D
8. Reconnaissance par Utilisation de Modèles 3D
9. Comparaison avec Des Travaux Antérieurs
10. Conclusions
  References

[1] T.D. Alter, and W.E.L. Grimson, Fast and robust 3d recognition by alignment. In Proc. IEEE International Conference on Computer vision, pp. 113-120, Berlin, Germany, 1993. 

[2] T.D. Alter and D.W. Jacobs, Error propagation in full 3d from 2d object recognition. In Proc. IEEE International Conference on Computer Vision and Pattern Recognition, pp. 892-898, Seatle, Washington, 1994. 

[3] N. Ayache, and O.D. Faugeras, A new approach for the recognition and positioning of two-Dimensional objects. IEEE Trans. Pattern Analysis and Machine Intelligence, 8(1): pp. 44-54, January 1986. 

[4] DH. Ballard, Generalized hough transform to detect arbitrary shapes, Pattern Recognition, 13(2): pp. 111-122, 1982. 

[5] J.S. Beis, and D.G. Lowe, Learning indexing functions for 3-d modelbased object recognition. In Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp. 275-280, Seatle, Washington, 1994. 

[6] J.R. Beveridge, and E.M. Riseman. Optimal geometric model matching under full 3d perspective. Computer Vision and Image Understanding, 61(3): pp. 351-364, 1995. 

[7] T.M. Breuel, Fast recognition using adaptive subdivisions of transformation Space,In Proc. IEEE International Conference on Computer vision and Pattern Recognition, pp. 445-451, Champain, Illinois, 1992. 

[8] T.M. Breuel, Geometric Aspects of Visual Object Recognition. PhD thesis, M.I.T., May 1992. 

[9] R. Brunelli, and D. Falavigna, Person identification using multiple cues. IEEE Trans. Pattern Analysis and Machine Intelligence, 17(10): pp. 955966, October 1995. 

[10] A.C. Califano, and R. Mohan, Multidimensional Indexing for Recognizing Visual Shapes. IEEE Trans. Pattern Analysis and Machine Intelligence, 16(4): pp. 373-392, April 1994. 

[11] T.A. Cass, Feature matching for object localization in the presence of uncertainty. In Proc. IEEE International Conference on Computer Vision, pp. 360-364, Osaka, Japan, 1990. 

[12] T.A. Cass, Polynomial-time geometric matching for object recognition. International Journal of Computer Vision, 21(1/2): pp. 37-61, 1997. 

[13] L.S. Davis, Hierarchical generalized hough transforms and line segment based generalized hough transforms. Pattern Recognition, 15(4): pp. 277285, 1982.

[14] D.F. DeMenthon, and L.S. Davis, Recognition and tracking of 3d objects by 1d search. In ARPA Image Understanding Workshop, pp. 653-659, Washington, DC, 1993. 

[15] R. Fletcher, Practical Methods of Optimization, John Wiley and Sons, New York, wiley-interscience publications edition, 1987. 

[16] T.L. Gandhi, and O.I. Camps, Robust feature selection for object recognition using uncertain 2d Image Data. In Proc. IEEE Conference on Computer Vision and pattern Recognition, pp. 281-287, Seatle, Washington, 1994. 

[17] D.M. Gavrila, and F.C.A. Groen, 3d object recognition from 2d images using geometric hashing. Pattern Recognition Letters, 13: pp. 263-278, 1992. 

[18] W.E.L. Grimson, The combinatorics of heuristic search term for object recognition in cluttered environment, IEEE Trans. Pattern Analysis and Machine Intelligence, 13 (9): pp. 920-935, 1991. 

[19] W.E.L. Grimson, and D.P. Huttenlocher, On the sensitivity of the hough transform for object recognition. IEEE Trans. Pattern Analysis and Machine Intelligence, 12(3) pp. 255-274, March, 1990.

[20] W.E.L. Grimson, and D.P. Huttenlocher, and D.W. Jacobs, A study of affine matching with bounded sensor error. In Proc. European Conference on Computer Vision, pp. 291-306, Santa Margherita Ligure, Italy, 1992. 

[21] W.E.L. Grimson, and T. Lozano-Perez, Localizing overlapping parts by searching the interpretation tree. IEEE Trans. Pattern Analysis and Machine Intelligence, 9(4): pp. 469-482, July, 1987. 

[22] Y. Hel-Or, and M. Werman, Absolute orientation from uncertain point data : A unified approach. In Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp. 77-82, Champain, Illinois, 1992. 

[23] D.P. Huttenlocher, and T.A. Cass, Measuring the quality of hypotheses in model-based recognition. In Proc. European Conference on Computer Vision, pp. 773-777, Santa Margherita Ligure, Italy, 1992. 

[24] D.P. Huttenlocher, and G.A. Klanderman, and W.J. Rucklidge, Comparing images using the hausdorff Distance. IEEE Trans. Pattern Analysis and Machine Intelligence, 15(9): pp. 850-863, 1993.

[25] D.P. Huttenlocher, and S. Ullman, Recognizing solid objects by alignment with an image. International Journal of Computer Vision, 5(2): pp. 195212, November 1990. 

[26] D.W. Jacobs, Robust and efficient detection of salient convex groups. IEEE Trans. Pattern Analysis and Machine Intelligence, 18(1): pp. 541548, 1996. 

[27] F. Jurie, Solution of the simultaneous pose and correspondence problem using gaussian error model. Computer Vision and Image Understanding, 73(3): pp. 357-373, 1999. 

[28] R. Kumar, and A.R. Hanson, Robust methods for estimating pose and a sensitivity analysis. Computer Vision and Graphics Image Processing, 60 (3): pp. 313-342, 1994. 

[29] D.G. Lowe, Robust model-based motion tracking through the integration of search and estimation. International Journal of Computer Vision, 8(2): pp. 113-122, 1992. 

[30] H. Murase and S.K. Nayar. Visual learning and recognition od 3d object from appearance. International Journal of Computer Vision, 18 (14): pp. 5-24, 1995. 

[31] C.F. Olson, Time and space efficient pose clustering. In Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp. 251-258, Seatle, Washington, 1994. 

[32] A.P. Pentland, and B. Moghadam, and  T. Starner, View-based and modular eigenspaces for face recognition. In Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp. 84-91, Seatle, Washington, 1994. 

[33] K.B. Sarachik, and W.E.L. Grimson, Gaussian error models for object recognition. IEEE Conference on Computer Vision and Pattern Recognition, pp. 400-406, New York, 1993. 

[34] S. Sclaroff, and A.P. Pentland, Modal matching for correspondence and recognition. IEEE Pattern Analysis and Machine Intelligence, 17: pp. 545-561, June 1995. 

[35] B.M. ter Haar Romeny, and L.M.J. Florack, and A.H. Salden, and M.A. Viergever, Higher order differential structure of images. Image and Vision Computing, 12(6): pp. 317-325, 1994. 

[36] W.M. Wells. Statistical approaches to feature-based object recognition. International Journal of Computer Vision, 21(1/2): pp. 63-98, 1997.