There is an enormous increase in the usage of an image databases across the world in the present digital scenario. In order to utilize such huge databases we are in need of an efficient image retrieval mechanism. By employing Content Based Medical Image Retrieval (CBMIR) system a far-reaching investigations has been carried out. In this research work a system named CBMIR is projected, which identifies the resemblance amongst the images by extraction of the image texture features, best features selection, categorization of the finest features. The feature which is picked is Robust Local Binary Patterns (RLBP); here every feature that is mined will be stored as a feature database. Fuzzy based Grey Wolf Optimization (FGWO) is employed to pick the most useful features and to decrease the texture features high dimensionality. To discover the best subset of features the evaluation criterion employed is a Classification algorithm. To classify the subset of images texture features a Fuzzy based C5.0 centered Decision tree is utilized. Performance metrics such as precision, recall and accuracy is employed for evaluation of the projected CBMIR system.
Volume 11 | 10-Special Issue
Pages: 676-681
DOI: 10.5373/JARDCS/V11SP10/20192857