A medical image identification system based on mixture models

Content Based Image Retrievals has become the most abbreviated thrust area today. The article we propose is a methodology for identifying the images based on relevancy using Kullback-Leibler method together with Generalized Gamma mixture model. The experimentation is carried out on the medical dataset namely med.univ-rennes1.fr and the results derived are compared for accuracy in terms of better perception. The results showcase that the performance of the method is about 84% and it is also performing efficiently in case of huge datasets.