Title :
Edge-region segmentation process based on generalized Voronoi diagram representation
Author :
Melkemi, M. ; Chassery, J.M.
Author_Institution :
Lab. TIM3-IMAG, CERMO, Grenoble, France
fDate :
30 Aug-3 Sep 1992
Abstract :
Cooperative methods provide comfortable solutions to segmentation problems. Different works have been engaged in this way in order to combine edge and region based methods. The authors propose a split and merge approach to decompose the image into homogeneous Voronoi polygons. The cooperative aspect comes from the fact that some contours of Voronoi polygons are supported by frontiers issued from an edge detector. From a geometrical point of view, the originality of this work is in the use of punctual Voronoi diagrams as an approximation of generalized Voronoi diagrams based on polygonal seed
Keywords :
computational geometry; edge detection; image segmentation; computational geometry; edge detector; edge-region segmentation; generalized Voronoi diagram representation; homogeneous Voronoi polygons; polygonal seed; punctual Voronoi diagrams; split and merge approach; Data structures; Detectors; Euclidean distance; Extremities; Image edge detection; Image segmentation; Microscopy; Partitioning algorithms; Reconnaissance; Sampling methods;
Conference_Titel :
Pattern Recognition, 1992. Vol.III. Conference C: Image, Speech and Signal Analysis, Proceedings., 11th IAPR International Conference on
Conference_Location :
The Hague
Print_ISBN :
0-8186-2920-7
DOI :
10.1109/ICPR.1992.201990