Level Set Formulation For Dual Snake Models
Dual snake models are powerful techniques for boundary extraction and segmentation of 2D images. In these methods one contour contracts from outside the target and another one expands from inside as a balanced technique with the ability to reject local minima. Such approaches have been proposed in the context of parametric snakes and extended for topologically adaptable snake models through the Dual-T-Snakes. In this chapter we present an implicit formulation for dual snakes based on the level set approach. The level set method consists of embedding the snake as the zero level set of a higher-dimensional function and to solve the corresponding equation of motion. The key idea of our work is to view the inner/outer contours as a level set of a suitable embedding function. The mathematical background of the method is explained and its utility for segmentation of cell images discussed in the experimental results. Theoretical aspects are considered and comparisons with parametric dual models presented.
KeywordsActive Contour Deformable Model Active Contour Model Gradient Vector Flow Medical Imaging Segmentation
Unable to display preview. Download preview PDF.
- 6.Black A, Yuille A, eds. 1993. Active vision. Cambridge: MIT Press.Google Scholar
- 10.Giraldi GA, Oliveira AF. 1999. Convexity analysis of snake models based on hamiltonian formu-lation. Technical report, Universidade Federal do Rio de Janeiro, Departamental Engenharia de Sistemas e Computa ç ão, http://www.arxiv.org/abs/cs.CV/0504031.
- 13.Giraldi GA,Vasconcelos N, Strauss E, OliveiraAF. 2001. Dual and topologically adaptable snakes and initialization of deformable models. Technical report, National Laboratory for Scientific Computation, Petropolis, Brazil, http://www.lncc.br/proj-pesq/relpesq-01.htm.
- 14.Bamford P, Lovell B. 1997. A two-stage scene segmentation scheme for the automatic collection of cervical cell images. In Proceedings of TENCON ’97: IEEE region 10 annual conference on speech and image technologies for computing and telecommunications, pp. 683-686.Washington, DC: IEEE.CrossRefGoogle Scholar
- 15.Giraldi GA, Strauss E, Oliveira AF. 2000. A boundary extraction method based on Dual-T-snakes and dynamic programming. In Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR ’2000), Vol. 1, pp. 44-49. Washington, DC: IEEE.Google Scholar
- 16.Giraldi GA, Strauss E, Oliveira AF. 2001. Improving the dual-t-snakes model. In Proceed-ings of the international symposium on computer graphics, image processing and vision (SIB-GRAPI’2001), Florian ópolis, Brazil, October 15-18, pp. 346-353.Google Scholar
- 18.Gunn SR. 1996. Dual active contour models for image feature extraction. Phd dissertation, Uni- versity of Southampton.Google Scholar
- 19.Bamford P, Lovell B. 1995. Robust cell nucleus segmentation using a viterbi search based active contour. In Proceedings of the fifth international conference on computer vision (ICCV’95), Cambridge, MA, USA pp. 840-845.Google Scholar
- 24.McInerney TJ. 1997. Topologically adaptable deformable models for medical image analysis. PhD dissertation. University of Toronto.Google Scholar
- 25.Sethian JA. 1996. Level set methods: evolving interfaces in geometry, fluid mechanics, computer vision and materials sciences. Cambridge: Cambridge UP.Google Scholar
- 29.Mauch S.2000. A fast algorithm for computing the closest point and distance function. Unpublished technical report. California Institute of Technology, September. http://www.acm.caltech.edu/ seanm/projects/cpt/cpt.pdf.
- 30.Sigg C, Peikert R. 2005. Optimized bounding polyhedra for GPU-based distance transform. In Scientific visualization: the visual extraction of knowledge from data, pp. 65-78. Ed GP Bonneau, T Ertl, GM Nielson. New York: Springer.Google Scholar
- 34.Pohle R, Behlau T, Toennies KD. 2003. Segmentation of 3D medical image data sets with a combination of region based initial segmentation and active surfaces. In Progress in biomedical optics and imaging. Proc SPIE Med Imaging 5203:1225-1231.Google Scholar
- 35.Falc ão AX, da Cunha BS, Lotufo RA. 2001. Design of connected operators using the image foresting transform. Proc SPIE Med Imaging 4322:468-479.Google Scholar
- 37.Xu C, Prince JL. 2000. Global optimality of gradient vector flow. In Proceedings of the 34th annual conference on information sciences and systems (CISS’00). iacl.ece.jhu.edu/pubs/p125c.ps.gz.Google Scholar
- 38.Hang X, Greenberg NL, Thomas JD. 2002. A geometric deformable model for echocardiographic image segmentation. Comput Cardiol 29:77-80.Google Scholar
- 39.Leventon ME, Eric W, Grimson L, Faugeras OD. 2000. Statistical shape influence in geodesic active contours. Proc IEEE Int Conf Pattern Recognit 1:316-323.Google Scholar
- 45.Cates JE, Lefohn AE, Whitaker RT. 2004. GIST: an interactive, GPU-based level set segmenta-tion tool for 3D medical images. http://www.cs.utah.edu/research/techreports/2004/pdf/UUCS-04-007.pdf.
- 46.Franca LP, Madureira AL, Valentin F. 2003. Towards multiscale functions: Enriching finite el-ement spaces with local but not bubble-like functions. Technical Report no. 26/2003, National Laboratory for Scientific Computing, Petropolis, Brazil.Google Scholar
- 48.Sapiro G.1995. Color snakes. Technical report, Hewlett-Packard Laboratories (email@example.com).Google Scholar