Title :
Unsupervised segmentation of textured satellite and aerial images with Bayesian methods
Author :
Wilson, Simon P. ; Zerubia, Josiane
Author_Institution :
Dept. of Stat., Trinity Coll., Dublin, Ireland
Abstract :
We investigate Bayesian solutions to unsupervised image segmentation based on the double Markov random field model. Inference on the number of classes in the image is done with reversible jump Metropolis moves. These moves are implemented by splitting and merging classes. Tests are conducted on satellite and aerial images.
Keywords :
Bayes methods; Markov processes; geophysical image processing; image segmentation; image texture; remote sensing; Bayesian method; Markov random field model; aerial images; merging class; satellite image texture; splitting class; unsupervised image segmentation; Abstracts; Bayes methods; Image segmentation; Markov processes;
Conference_Titel :
Signal Processing Conference, 2002 11th European
Conference_Location :
Toulouse
