Detection of Regular Boundaries in Noisy and Textured Images a Multiscale Active Contour Approach. Détection de Contours Réguliers dans des Images Bruitées et Texturées une Approche par Contours Actifs Multiéchelle

Detection of Regular Boundaries in Noisy and Textured Images a Multiscale Active Contour Approach

Détection de Contours Réguliers dans des Images Bruitées et Texturées une Approche par Contours Actifs Multiéchelle

Pierre-Louis Bossart Dominique David  Jean-Marc Dinten  Jean-Marc Chassery 

Eli (CEA- Technologies Avancées), Département Systèmes 17, Avenue desMartyrs, F-38054 Grenoble Cedex 9

Laboratoire TIMC-IMAG Institut Albert Bonniot Domaine de la Merci F-38706 La Tronche

Page: 
209-225
|
Received: 
5 January 1995
| |
Accepted: 
N/A
| | Citation

OPEN ACCESS

Abstract: 

This article deals with the detection of edges in noisy and textured images . After an overview of the shortcomings of local operators, we assess active contour models, which constrain the detection using a priori information . The implementation of active contours is difficult due to several problems. We suggest two methods which make this global approach more robust and easier to use. On one hand, the parameters are set using a calibration algorithm which relies on geometrical a priori . The value of the parameters depends upon the maximum curvature . On the other hand, a multiscale strategy reduces the need for an accurate initialization of active contours. The convergence is robust at coarse scales and the localization of edges is enhanced at fine scales. Several efficient algorithms are proposed to track contours over decreasing scale, using a prediction on the distorsion of boundaries caused by gaussian blurring. The experiments show the validity of our approach and the relevance of a cooperation between segmentation processes, especially to solve the initialization problem.

Résumé 

Cet article traite de la détection de contours dans des images fortement bruitées et texturées . Après avoir exposé les limitations des opérateurs locaux, nous suggérons de contraindre la détection grâce aux modèles de contours actifs, qui introduisent des informations a priori sur la géométrie et la régularité des objets cherchés. La mise en oeuvre des contours actifs est difficile en raison de nombreux problèmes pratiques . Nous proposons deux techniques rendant cette approche globale plus robuste et plus facile d'emploi. D'une part, nous facilitons le choix des paramètres en adaptant un algorithme de calibrage ayant une explication géométrique : les paramètres dépendent de la valeur de la courbure maximale . L'association des contours actifs et d'une représentation multiéchelle permet d'autre part de réduire la dépendance vis-à-vis de l'initialisation . Après une convergence robuste vers une solution grossière, la localisation des contours est améliorée en diminuant progressivement l'échelle d'analyse. Nous proposons alors plusieurs techniques efficaces de suivi des contours dans l'espace-échelle, s'appuyant sur une prédiction du déplacement des frontières sous l'effet du lissage gaussien . Les résultats expérimentaux montrent la validité de notre approche, et mettent en évidence l'apport d'une collaboration entre processus de segmentation, en particulier pour automatiser l'initialisation.

Keywords: 

Edge detection, Active contours, Calibration of parameters, Multiscale representation, Scale-space, Edge focusing.

Mots clés 

Détection de contours, Contours actifs, Calibrage des paramètres, Représentations multiéchelles, Espace-échelle, Suivi de contours.

1. Introduction
2. Le Modèle Classique de Contour Actif
3. Calibrage des Paramètres
4. Formulation des Forces Externes et Calibrage
5. Construction d'une Représentation Multiéchelle
6. Suivi de Contours
7. Résultats Expérimentaux
8. Conclusion
  References

[1] J. Canny. A computational approach to edge detection. I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 8(6) :679-697, November 1986. 

[2] J. Shen and S. Castan. An optimal linear operator for edge detection. In Proceedings of the Conference on Computer Vision and Pattern Recognition, Miami, FL, pages 109-114, June 1986. 

[3] R. Deriche. Using Canny's criteria to derive a recursively implemented optimal edgedetector.International Journal of Computer Vision, 1(2) 

[4] L. Vincent.Algorithmes morphologiques à base de files d'attenteet de lacets Extension aux graphes. Thèse,Ecoledes Mines de Paris, 1990. 

[5] S. Geman and D. Geman. Stochastic relaxation, Gibbs distributions and the bayesian restoration of images.I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 6(6) :721-741, November 1984. 

[6] D. Geiger and A. Yuille. A common framework for image segmentation. International Journal of Computer Vision, 6(3) :227-243, 1991. 

[7] L.H. Stainand J.S. Duncan. Boundary finding with parametrically deformable models.I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 1400 :1061-1075, November 1992. 

[8] U. Grenander, Y. Chow, and D.M. Keenan. HANDS, A Pattern Theoretical Study of Biological Shapes. Springer-Verlag, New-York, 1991. 

[9] A.L. Yuille, P. W. Hallinan, and D.S. Cohen. Feature extraction from faces using deformable templates. International Journal ofComputer Vision, 8(2) :99111,1992.

[10] F. Solina and R. Bajcsy. Recovery of parametric models from range images The case for superquadrics. I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 12(2) :131-147, February 1990. 

[11] M. Kass, A. Witkin, and D. Terzopoulos. Snakes : Active contour models. International Journal of Computer Vision, 1 :312-331, 1988. 

[12] 1. Cohen, L.D. Cohen, and N. Ayache. Using deformable surfaces to segment 3d images and infer differential structure. CVGIP- Image Understanding, 56(2) :242-263, September 1992. 

[13] S. Menet, P. SaintMarc, and G. Medioni. Active contour models :Overview, implementation and applications. InProceedings of the I.E.E.E. Conference on Systems, Man and Cybernetics,pages 194-199, 1990. 

[14] R. Cipolla and A. Blake. Surface shape from the deformations of apparent contours. International Journal of Computer Vision, 9(2) :83-112, 1992. 

[15] P-L. Bossart. Détection de contours réguliers dans des images bruitées et texturées : association des contours actifs et d'une approche multiéchelle. Thèse, InstitutNational Polytechnique deGrenoble, Octobre 1994. 

[16] M-O. Berger. Les contours actifs : Modélisation, comportement, convergence.Thèse, Institut NationalPolytechniquedeLorraine, Nancy, 1991. 

[17] N. Rougon and F.Prêteux. Régularisation directionnelle optimale etmodèles déformables adaptatifs. In Actesdu CongrèsReconnaisance des Formes et IntelligenceArtificielle, pages 51-62. AFCET, Janvier 1994. 

[18] L.D. Cohen. On active contour models and balloons. CVGIP- Image Understanding, 53(2) :211-218, March 1991. 

[19] D. Terzopoulos. On matching deformable models to images. In Topical Meeting on Machine Vision (Optical Society ofAmerica), Technical Digest Series,pages 160-163, Incline Village, Nevada, USA, March 1987. 

[20] F. Leymarie and M.D. Levine. Tracking deformable objects in the plane using an active contour model. IEEE. Transactions on Pattern Analysis and Machine Intelligence, 15(6) :617-634, June 1993.

[21] J-M. Dinten and A. Trouvé. A deformable model approach for the determination of transition strips on radiographic images . InProceedingsofthe 11th International Conference on Pattern Recognition, The Hague, The Netherlands,pages 355-358, 1992. 

[22] P. Fua and YG. Leclerc.Model driven edge detection. Machine Vision and Applications, 3 :45-56, 1990. 

[23] C.A. Davatzikos and J.L. Prince. Adaptive active contour algorithms for extracting and mapping thick curves. In Proceedings ofthe Computer Vision and Pattern Recognition Conference, New York, pages 524-529, 1993. 

[24] L.D. Cohen, E. Bardinet, and N. Ayache. Reconstruction of digital terrain model with a lake. InProceedings of theSPIEVol. 2031 Geometric Methods in Computer Vision II, pages 38-50, 1993. 

[25] T. McInerney and D. Terzopoulos. A finite element model for 3D shape reconstruction and nonrigid motion tracking. In Proceedings of the IEEE International Conference on Computer Vision, pages 518-523, 1993. 

[26] L.D. Cohen, E. Bardinet, and N. Ayache. Reconstruction of digital terrain model with a lake. InProceedings of theSPIE Vol. 2031 Geometric Methods in Computer Vision 11, pages 38-50, 1993.

[27] H. Delinguette ans M. Hebert and K. Ikeuchi. Shape representation and image segmentation using deformable surfaces. InProceedings ofthe IEEE International Conference on Computer Vision and Pattern Recognition, pages 467-472,1991. 

[28] D. Mart. Vision. W.H. Freeman, San Francisco, CA, 1982. 

[29] P.J. Burt. The laplacian pyramid as a compact image code. I.E.E.E. Transactions on Communications, 31(4) :532-540, April 1984.

[30] J.L. Crowley and R.M. Stern. Fast computation of the difference of lowpass transform. IEE.E. Transactions on Pattern Analysis and Machine Intelligence, 6(2):212-222, March 1984. 

[31] A. Chéhikian. Algorithmes optimaux pour la génération de pyramides d'images passe-bas et laplaciennes. Traitement du Signal, 9(4) :297-307, 1992. 

[32] S. Mallat. A theory for multiresolution signal decomposition :The wavelet representation.LE.E.E. Transactions on Pattern Analysis and Machine Intelligence. 11(7) :674-693, July 1989. 

[33] I. Daubechies. Ten Lectures on Wavelets, volume 61 of CBMS Regional Conference Series in Applied Mathematics .Society for Industrial and Applied Mathematics, Philadelphia, PA, 1992. 

[34] A. Witkin. Scale-space filtering. In Proceedings of the International Joint Conference on Artificial Intelligence, Karlsruhe, pages 1019-1021, 1983. 

[35] P. Perona and J. Malik. Scale-space and edge detection using anisotropic diffusion.I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 12(7) :629-639, July 1990. 

[36] J-M. Morel, P-L. Lions, and L. Alvarez. Image selective smoothing and edge detection by non-linear diffusion. Preprint, Université de Paris-Dauphine, 1991. 

[37] T. Lindeberg. Scale-space for discrete signals. IEEE. Transactions on Pattern Analysis and Machine Intelligence, 12(3) :234-254, March 1989. 

[38] V. Berzins. Accuracy of Laplacian edge detectors.Computer Vision, Graphics and Image Processing, 27 :195-210, 1984. 

[39] F. Bergholm.Edge, focusing. I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 9(6) :726-74 1, November 1987. 

[40] Y.Luand R.C. Jain. Behavior of edges in scale-space.I.E.E.E. Transactions on Pattern Analysis and Machine Intelligence, 11(4) :337-356, April 1989. 

[41] R. Deriche and G. Giraudon. A computational approach for corner and vertex detection. International Journal ofComputer Vision, 10(2) :101-124, 1993. 

[42] Y . Elomary. Modèles déformables et multirésolution pour la détection de contours entraitement d'images. Thèse, UniversitéJoseph Fourier, Grenoble, 1994. 

[43] Y. Lu and R.C. Jain. Reasoning about edges in scale space . LE.E.E. Transactions on Pattern Analysis and Machine Intelligence, 14(4) :450-468, April 1992. 

[44] V. Caselles, F. Catte, B. Coll, and F. Dibos. A geometric model for active contours in image processing.Cahiers du CEREMADE, Universitéde ParisDauphine, 1992. 

[45] R. Malladi, J.A. Sethian, and B.C. Vemuri. Shape modeling with front propagation: A level set approach.I.E.EE . Transactions on Pattern Analysis and Machine Intelligence, 17(2) :158-175, February 1995. 

[46] I. Herlin, C. Nguyen, and C. Graffigne. A deformable region model using stochastic processes applied to echocardiographic images. InProceedings of the Conference on Computer Vision and Pattern Recognition, 1992. 

[47] P-L. Bossart.Hypertoolsin image and volume visualization. InProceedings ofthe Tel'96Workshop, Monterrey, CA.USENIX, July 10-13 1996. Available at ftp ://redhook.llnl.gov/pub/visu/papers.

[48] P-L. Bossart, H.E. Martz, and K. Hollerbach. Finite element analysis of human joints. In Proceedings of the ICIP'96 Conference, Lausanne, Switzerland. I.E.E.E., September 16-19 1996. Available at ftp ://redhook.llnl.gov/pub/visu/papers.