Sistema para segmentaÃÃo de imagens por agrupamento hierÃrquico baseado no comportamento superparamagnÃtico do modelo de Potts




Recently an analogy was established between the problem of clustering data and the search of typical configurations of a physical model, the Potts model. To illustrate this analogy, a program was developed, in the C language, by Professor Eytan Domany (Department of Physics of Complex Systems, Weizmann Institute of Science, Israel). The main purpose of this program is the accomplishment of a hierarchical clustering algorithm based on the superparamagnetic behavior of the Potts model (SPC - Superparamagnetic Clustering), through finding typical states of a non-homogeneous Potts model in intermediate states between the ferromagnetic and the paramagnetic phases. This work is aimed at clustering technique to image segmentation. An interface, using IDL and ENVI, was developed in order to assist the user in the use of this and other segmentation procedures and to provide tools to understand the results from hierarchical clustering techniques. Clusters are illustrated through images, interactive hierarchical graphs (dendrogram) and animations


interface interface agrupamento hierÃrquico de dados, modelo de potts ciencia da computacao hierarchical clustering data segmentaÃÃo de imagens potts model image segmentation

Documentos Relacionados