A fuzzy clustering algorithm with the criterion of modified objective function for application of automatic target recognition systems

Structure organization of the physical geometry in real data is significantly indicated by characteristics of embodied pattern classifiers. The requirement that this paper considers for resolving this issue is a suitable measure of clusters with the criterion of objective functionals. It involves the optimizing of partition coefficient and entropy and separation coefficients through implementation of an MFCM algorithm to evaluate the clustering. In this way a clustering experiment for recognition of an ambiguous photograph of a MIG-29 fighter taken by spy satellite has been successfully developed by computer calculation.