New Elastica Geodesic Approach with Convexity Shape Prior for Region-based Active Contours and Image Segmentation
Chen, Da; Cohen, Laurent D.; Mirebeau, Jean-Marie; Tai, Xue-Cheng (2021), New Elastica Geodesic Approach with Convexity Shape Prior for Region-based Active Contours and Image Segmentation, ICCV 21, International Conference on Computer VIsion, 2021-10, Montreal, CANADA
TypeCommunication / Conférence
Lien vers un document non conservé dans cette basehttps://hal.archives-ouvertes.fr/hal-03174123
Titre du colloqueICCV 21, International Conference on Computer VIsion
Date du colloque2021-10
Ville du colloqueMontreal
Pays du colloqueCANADA
MétadonnéesAfficher la notice complète
Résumé (EN)The minimal geodesic models based on the Eikonal equations are capable of finding suitable solutions in various image segmentation scenarios. Currently, existing geodesic-based segmentation approaches usually exploit the image features in conjunction with regularization terms, such as curve length, for computing geodesic paths. In this paper, we consider a more complicated problem: finding simple closed geodesic curves which are imposed a convexity shape prior. The proposed approach relies on an orientation-lifting strategy, by which a planar curve can be mapped to an high-dimensional orientation space. The convexity shape priors serve as a constraint for the construction of local metrics in the lifted space. The geodesic curves then can be efficiently computed through the single-pass Fast Marching method (FMM). In addition, we introduce a way to incorporate region-based homogeneity features into the proposed geodesic model so as to solve the region-based segmentation issues with shape prior constraints.
Mots-clésActive contours; Convexity shape prior; Curvature penalization; Eikonal equation
Affichage des éléments liés par titre et auteur.