Title :
Modeling noise for a better simplification of skeletons
Author :
Attali, Dominique ; Montanvert, Annick
Author_Institution :
TIMC-IMAG, La Tronche, France
Abstract :
The skeleton of an object is the locus of the centers of maximal discs included in the shape. The skeleton provides a compact representation of objects, useful for shape description and recognition. A well-known drawback of the skeleton transformation is its lack of continuity. This paper is concerned with the modeling of noise that may affect objects and the consequence of this noise on the skeleton. A graph (called the parameter graph) is introduced, on which branches due to noise are characterized. We deduce from this preliminary study a method to simplify skeletons. It depends on thresholds that can be chosen directly on the parameter graph associated to each skeleton
Keywords :
graph theory; image recognition; image representation; mathematical morphology; noise; compact representation; continuity; maximal discs; modeling noise; object skeleton; parameter graph; shape description; shape recognition; skeleton transformation; skeletons simplification; Discrete transforms; Geometry; Image analysis; Image reconstruction; Laboratories; Noise shaping; Shape measurement; Skeleton; Vocabulary;
Conference_Titel :
Image Processing, 1996. Proceedings., International Conference on
Conference_Location :
Lausanne
Print_ISBN :
0-7803-3259-8
DOI :
10.1109/ICIP.1996.560357