Evaluating a Cover Based Rough Set Classifier in a Content Based Image Retrieval System

One of the more important computer vision tasks is content based image retrieval system. Classifiers are playing an important role in content based image retrieval systems. To address low retrieval accuracy rates in classifiers, this study focused on applying cover based rough set algorithm to build a rule based classifier for content based image retrieval systems. Using a cover based rough set classifier, which can work efficiently with uncertainly and vagueness, can improve the overall accuracy of content based image retrieval systems. The execution of the cover based rough set classification is compared with other classifiers, such as Rough Set, Navies Byes, Support Vector Machine, Radial Basis Function Network and Decision Tree, using different evaluating measures in the Corel image dataset. The results of the experiments showed that using the cover based rough set classifier technique lead to different results and some of them can perform better than current well-known classifiers.