[Un algorithme de recalage rapide d'images]
On propose un algorithme permettant d'estimer un mouvement régulier de façon rapide. Cet algorithme est basé sur l'hypothèse de préservation du niveau de gris. Une approche hiérarchique permet de réduire l'espace des déplacements admissibles, et le nombre de paramètres est faible devant la taille des données. La fonction coût non-linéaire donnant le déplacement optimal est minimisée par la méthode de Gauss–Newton, et la matrice nécessaire à chaque pas est assemblée efficacement à l'aide d'une méthode d'éléments finis.
An algorithm is presented here to estimate a smooth motion at a high frame rate. It is derived from the non-linear constant brightness assumption. A hierarchical approach reduces the dimension of the space of admissible displacements, hence the number of unknown parameters is small compared to the size of the data. The optimal displacement is estimated by a Gauss–Newton method, and the matrix required at each step is assembled rapidly using a finite-element method.
Accepté le :
Publié le :
@article{CRMATH_2008__346_9-10_593_0,
author = {Fehrenbach, J\'er\^ome and Masmoudi, Mohamed},
title = {A fast algorithm for image registration},
journal = {Comptes Rendus. Math\'ematique},
pages = {593--598},
publisher = {Elsevier},
volume = {346},
number = {9-10},
year = {2008},
doi = {10.1016/j.crma.2008.03.019},
language = {en},
url = {http://www.numdam.org/articles/10.1016/j.crma.2008.03.019/}
}
TY - JOUR AU - Fehrenbach, Jérôme AU - Masmoudi, Mohamed TI - A fast algorithm for image registration JO - Comptes Rendus. Mathématique PY - 2008 SP - 593 EP - 598 VL - 346 IS - 9-10 PB - Elsevier UR - http://www.numdam.org/articles/10.1016/j.crma.2008.03.019/ DO - 10.1016/j.crma.2008.03.019 LA - en ID - CRMATH_2008__346_9-10_593_0 ER -
%0 Journal Article %A Fehrenbach, Jérôme %A Masmoudi, Mohamed %T A fast algorithm for image registration %J Comptes Rendus. Mathématique %D 2008 %P 593-598 %V 346 %N 9-10 %I Elsevier %U http://www.numdam.org/articles/10.1016/j.crma.2008.03.019/ %R 10.1016/j.crma.2008.03.019 %G en %F CRMATH_2008__346_9-10_593_0
Fehrenbach, Jérôme; Masmoudi, Mohamed. A fast algorithm for image registration. Comptes Rendus. Mathématique, Tome 346 (2008) no. 9-10, pp. 593-598. doi : 10.1016/j.crma.2008.03.019. http://www.numdam.org/articles/10.1016/j.crma.2008.03.019/
[1] A computational framework and an algorithm for the measurement of visual motion, Int. J. Comput. Vis., Volume 2 (1989), pp. 283-310
[2] The computation of optical flow, ACM Computing Surveys, Volume 27 (1995), pp. 433-467
[3] Computing large deformations metric mappings via geodesic flows of diffeomorphisms, Int. J. Comput. Vis., Volume 61 (2005) no. 2, pp. 139-157
[4] Determining optical flow, Artificial Intelligence, Volume 17 (1981), pp. 185-204
[5] B. Lucas, T. Kanade, An iterative image registration technique with an application to stereo vision, in: Proc. Seventh International Joint Conference on Artificial Intelligence, Vancouver, Canada, 1981, pp. 674–679
[6] Dense estimation and object-based segmentation of the optical flow with robust techniques, IEEE Trans. Image Proc., Volume 7 (1998), pp. 703-719
[7] A new image registration technique with free boundary constraints: application to mammography, Computer Vision and Image Understanding, Volume 89 (2003), pp. 166-196
Cité par Sources :
