A Novel and Multi-Scale Unsupervised Algorithm for Image Segmentation
Minmin Luo, Guiping Jiang, Yazhong Lin · 2010
Gibbs Random Fields (GRF) is a popular prior model widely used in Bayesian segmentation due to its excellent property in describing the spatial information of image. But until now, the classical approaches, describing the Markovian property of single-scale instead that of multi-scale, may come across some difficulties such as expensive computation and unsupervised parameter estimation of GRF. Thus, in this paper, a novel and unsupervised algorithm named multi-scale GRF that addresses these problems perfectly is proposed by extending the classical single-scale model of GRF to a multi-scale one at the first time. Experiments have shown that our algorithm presented in the paper has excellent robustness and easy to be used in unsupervised and precise segmentation.