[Gradient topologique pour des EDP du quatrième ordre et application à la détection de structures fines dans des images 2D]
Dans cette note, on décrit une nouvelle approche pour la détection de structures fines dans une image. Cette approche est basée sur le calcul du gradient topologique associé à une fonction coût définie à partir des dérivées secondes d'une régularisation des données (éventuellement bruitées). Cette régularisation est obtenue via la résolution d'une EDP du quatrième ordre. L'étude de la sensibilité topologique est faite dans les cas d'une inclusion circulaire et d'un crack. Nous illustrons notre approche en donnant deux résultats expérimentaux.
In this paper we describe a new approach for the detection of fine structures in an image. This approach is based on the computation of the topological gradient associated with a cost function defined from a regularization of the data (possibly noisy). We get this approximation by solving a fourth-order PDE. The study of the topological sensitivity is made in the cases of both a circular inclusion and a crack. We illustrate our approach by giving two experimental results.
Accepté le :
Publié le :
@article{CRMATH_2014__352_7-8_609_0, author = {Aubert, Gilles and Drogoul, Audric}, title = {Topological gradient for fourth-order {PDE} and application to the detection of fine structures in {2D} images}, journal = {Comptes Rendus. Math\'ematique}, pages = {609--613}, publisher = {Elsevier}, volume = {352}, number = {7-8}, year = {2014}, doi = {10.1016/j.crma.2014.06.005}, language = {en}, url = {http://www.numdam.org/articles/10.1016/j.crma.2014.06.005/} }
TY - JOUR AU - Aubert, Gilles AU - Drogoul, Audric TI - Topological gradient for fourth-order PDE and application to the detection of fine structures in 2D images JO - Comptes Rendus. Mathématique PY - 2014 SP - 609 EP - 613 VL - 352 IS - 7-8 PB - Elsevier UR - http://www.numdam.org/articles/10.1016/j.crma.2014.06.005/ DO - 10.1016/j.crma.2014.06.005 LA - en ID - CRMATH_2014__352_7-8_609_0 ER -
%0 Journal Article %A Aubert, Gilles %A Drogoul, Audric %T Topological gradient for fourth-order PDE and application to the detection of fine structures in 2D images %J Comptes Rendus. Mathématique %D 2014 %P 609-613 %V 352 %N 7-8 %I Elsevier %U http://www.numdam.org/articles/10.1016/j.crma.2014.06.005/ %R 10.1016/j.crma.2014.06.005 %G en %F CRMATH_2014__352_7-8_609_0
Aubert, Gilles; Drogoul, Audric. Topological gradient for fourth-order PDE and application to the detection of fine structures in 2D images. Comptes Rendus. Mathématique, Tome 352 (2014) no. 7-8, pp. 609-613. doi : 10.1016/j.crma.2014.06.005. http://www.numdam.org/articles/10.1016/j.crma.2014.06.005/
[1] Topological asymptotic analysis of the Kirchhoff plate bending problem, ESAIM Control Optim. Calc. Var., Volume 17 (2011) no. 3, pp. 705-721
[2] Crack detection by the topological gradient method, Control Cybern., Volume 34 (2005) no. 1, pp. 81-101
[3] Image restoration and classification by topological asymptotic expansion (Taroco, E.; de Souza Neto, E.A.; Novotny, A.A., eds.), Variational Formulations in Mechanics: Theory and Applications, CIMNE, Barcelona, Spain, 2007, pp. 23-42
[4] The Mathematical Theory of Finite Element Methods, Springer-Verlag, 2002
[5] Vesselness enhancement diffusion, Pattern Recognit. Lett., Volume 24 (2003) no. 16, pp. 3141-3151
[6] A. Drogoul, G. Aubert, Topological gradient for fourth order PDE and application to the detection of fine structures in 2D images, Control Optim. Calc. Var., submitted for publication.
[7] Application of the topological gradient to image restoration and edge detection, Eng. Anal. Bound. Elem., Volume 32 (2008) no. 11, pp. 891-899
[8] Nonconforming tetrahedral finite elements for fourth order elliptic equations, Math. Comput., Volume 76 (2007), pp. 1-18
[9] On the topological derivative in shape optimization, SIAM J. Control Optim., Volume 37 (1999) no. 4, pp. 1251-1272
[10] An unbiased detector of curvilinear structures, IEEE Trans. Pattern Anal. Mach. Intell., Volume 20 (1998) no. 2, pp. 113-125
Cité par Sources :