%0 Conference Proceedings
%T Image Segmentation by Image Foresting Transform with Geodesic Band Constraints
%D 2017
%A de Moraes Braz, Caio,
%A Vechiatto de Miranda, Paulo André,
%@affiliation IME-USP
%@affiliation IME-USP
%E Torchelsen, Rafael Piccin,
%E Nascimento, Erickson Rangel do,
%E Panozzo, Daniele,
%E Liu, Zicheng,
%E Farias, Mylène,
%E Viera, Thales,
%E Sacht, Leonardo,
%E Ferreira, Nivan,
%E Comba, João Luiz Dihl,
%E Hirata, Nina,
%E Schiavon Porto, Marcelo,
%E Vital, Creto,
%E Pagot, Christian Azambuja,
%E Petronetto, Fabiano,
%E Clua, Esteban,
%E Cardeal, Flávio,
%B Conference on Graphics, Patterns and Images, 30 (SIBGRAPI)
%C Niterói, RJ
%8 Oct. 17-20, 2017
%S Proceedings
%I Sociedade Brasileira de Computação
%J Porto Alegre
%K Image Segmentation, IFT, shape constraints.
%X The Image Foresting Transform framework (IFT) was successfully used to implement several segmentation meth- ods, including watersheds and fuzzy connectedness, however the lack of regularization terms in its formulation leads to a potential irregular (jagged) segmentation. An attempt to avoid this issue is to employ shape constraints that favor more regular shapes. We present a novel shape constraint, called the Geodesic Band Constraint (GBC) and show how it can be efficiently incorporated in the Image Foresting Transform framework, with its proof of optimality in terms of an energy function, subject to the new constraint. This constraint helps us to improve the segmentation of regular objects. The GBC can be also used with a prior shape template in order to drive the segmentation towards a specific shape with a single parameter that controls the degrees of freedom of the allowed deformations subject to the model.
%@language en
%3 CameraReady_CaioBraz.pdf