INTUITIONISTIC ROBUST CLUSTERING FOR SEGMENTATION OF LESIONS IN DERMATOSCOPIC IMAGES

Abstract

This paper presents the formulation of the intuitive fuzzy 
clustering algorithm to be robust to atypical data present in 
dermoscopic images and to delimit the affected area. This 
algorithm is formulated from the objective function derivation 
for memberships update, to integrate an m-redescending 
estimator influence function. Experimentation shows an 
accuracy of 95% with the proposal algorithm with respect to 
other clustering algorithms to perform delimitations, in 
addition the iterations number is considerably reduced.