Graph Cut Based Segmentation of Predefined Shapes: Applications to Biological Imaging

Emmanuel Soubies (1) , Pierre Weiss (2) , Xavier Descombes (1)

1 MORPHEME - Morphologie et Images
2 IMT - Institut de Mathématiques de Toulouse UMR5219

We propose an algorithm to segment 2D ellipses or 3D ellipsoids. This problem is of fundamental importance in various applications of cell biology. The algorithm consists of minimizing a contrast invariant energy defined on sets of non overlapping ellipsoids. This highly non convex problem is solved by combining a stochastic approach based on marked point processes and a graph-cut algorithm that selects the best admissible configuration. In order to accelerate the computing times, we delineate fast algorithms to assess whether two ellispoids intersect or not and various heuristics to improve the convergence rate.

Mots clés: Ellipses and ellipsoids; Multiple objects detection; Bio-imaging; Nuclei segmentation; 2D and 3D images; Graph-cuts; Multiple birth and cut; Marked point processes

Domaines: Traitement des images [eess.IV]

Emmanuel Soubies, Pierre Weiss, Xavier Descombes. Graph Cut Based Segmentation of Predefined Shapes: Applications to Biological Imaging. Pattern Recognition Applications and Methods., 318, Springer-Verlag, pp.153-170, 2015, Advances in Intelligent Systems and Computing, 978-3-319-12610-4. ⟨10.1007/978-3-319-12610-4_10⟩. ⟨hal-01088384⟩