Title :
Non-supervised segmentation using multi-level Markov random fields
Author :
Azencott, R. ; Graffigne, C.
Author_Institution :
Universite. de Paris-Sud, Orsay, France
fDate :
30 Aug-3 Sep 1992
Abstract :
Presents a region growing algorithm based on multi-level Markov random fields. The lower level is pixel-based and the higher ones are region-based, the region mean size increasing with the complexity level. The higher level algorithm has the particularity to allow the use of global geometric region information
Keywords :
Markov processes; image segmentation; global geometric region information; image segmentation; multi-level Markov random fields; pixel-based; region growing algorithm; region-based; unsupervised segmentation; Feeds; Image segmentation; Layout; Markov random fields; Merging; Multispectral imaging; Nearest neighbor searches; Noise level; Pattern recognition; Satellites;
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.1992.201961
