Image color segmentation using the fuzzy tree algorithm T-LAMDA

The image segmentation is very sensitive to the features used in the similarity measure and the objects type. In this paper we introduce a new segmentation algorithm based on fuzzy clustering. This method allows to incorporate spatial information which yield the result more accurate and more robust to noise. It is completely automatized with respect to the number of clusters and the setting up of membership functions. The data structure based on a Fuzzy Tree Algorithm allows to reduce the CPU time.