The Nonlinear Tensor Diffusion in Segmentation of Meaningful Biological Structures from Image Sequences of Zebrafish Embryogenesis - Université Pierre et Marie Curie Accéder directement au contenu
Chapitre D'ouvrage Année : 2009

The Nonlinear Tensor Diffusion in Segmentation of Meaningful Biological Structures from Image Sequences of Zebrafish Embryogenesis

Résumé

In this contribution we develop a strategy for segmentation of evolving biological structures in image sequences. Our approach is based on combination of nonlinear tensor diffusion image smoothing and subjective surface based image segmentation. Since the fine cell structure would restrain the evolving segmentation function to achieve a shape of meaningful biological structures, we have to smooth properly the images in the sequence. To that goal we apply the nonlinear tensor diffusion which enhances the connectivity of bordering structure lines and smoothes their inner parts. For the numerical implementations we use semi-implicit diamond-cell finite volume methods both for filtering and segmentation. We show application of the method in image segmentation of early stages of zebrafish embryogenesis.

Dates et versions

hal-00723371 , version 1 (09-08-2012)

Identifiants

Citer

Olga Drblikova-Stasova, Karol Mikula, Nadine Peyriéras. The Nonlinear Tensor Diffusion in Segmentation of Meaningful Biological Structures from Image Sequences of Zebrafish Embryogenesis. Scale Space and Variational Methods in Computer Vision, Springer Verlag, pp.63-74, 2009, Lecture Notes in Computer Science Volume 5567, 9783642022555. ⟨10.1007/978-3-642-02256-2_6⟩. ⟨hal-00723371⟩
112 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More