Interactive segmentation based on component-trees

Benoît Naegel (1) , Nicolas Passat (2)

1 ICube - Laboratoire des sciences de l'ingénieur, de l'informatique et de l'imagerie
2 CRESTIC - Centre de Recherche en Sciences et Technologies de l'Information et de la Communication - EA 3804

Component-trees associate to a discrete gray-level image a descriptive data structure induced by the inclusion relation between the binary components obtained at successive level-sets. This article presents an interactive segmentation methodology based on component-trees. It consists of the extraction of a subset of the image component-tree, enabling the generation of a binary object which fits at best (with respect to the gray-level structure of the image) a given binary target selected beforehand in the image. Compared to other interactive segmentation methods, the proposed methodology has the following advantages: (i) the segmentation result is only composed of a union of connected components of the level-sets, which ensures that no "false contours" are included; (ii) only one image marker is needed: in particular, there is no need to give a marker for the background (contrary to some other methods); (iii) the method is fast and efficient, leading to a result computed in real-time on common image sizes.

Mots clés: mathematical morphology; segmentation; component-tree

Domaines: Traitement des images [eess.IV]; Vision par ordinateur et reconnaissance de formes [cs.CV]

Benoît Naegel, Nicolas Passat. Interactive segmentation based on component-trees. Image Processing On Line, 2014, 4, pp.89-97. ⟨10.5201/ipol.2014.71⟩. ⟨hal-01694403⟩