Unsupervised Hierarchical Image Segmentation based on the TS-MRF model and Fast Mean-Shift Clustering

R. Gaetano (1, 2) , G. Scarpa (2) , G. Poggi (2) , J. Zerubia (1)

1 ARIANA - Inverse problems in earth monitoring
2 DIET - Dipartimento di Ingegneria Elettronica e delle Telecomunicazioni

Tree-Structured Markov Random Field (TS-MRF) models have been recently proposed to provide a hierarchical multiscale description of images. Based on such a model, the unsupervised image segmentation is carried out by means of a sequence of nested class splits, where each class is modeled as a local binary MRF. We propose here a new TS-MRF unsupervised segmentation technique which improves upon the original algorithm by selecting a better tree structure and eliminating spurious classes. Such results are obtained by using the Mean-Shift procedure to estimate the number of pdf modes at each node (thus allowing for a non-binary tree), and to obtain a more reliable initial clustering for subsequent MRF optimization. To this end, we devise a new reliable and fast clustering algorithm based on the Mean-Shift technique. Experimental results prove the potential of the proposed method.

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

R. Gaetano, G. Scarpa, G. Poggi, J. Zerubia. Unsupervised Hierarchical Image Segmentation based on the TS-MRF model and Fast Mean-Shift Clustering. Proc. European Signal Processing Conference, EUSIPCO 2008, Aug 2008, Lausanne (CH), Switzerland. ⟨inria-00503198⟩