A Novel Based Decision Tree for Content Based Image Retrieval: An Optimal Classification Approach

In this 21
^st
century, the development of multimedia technology is at its peak. Content based image retrieval (CBIR) is an Inventive methodology which can retrieve images based on consequently inferred highlights, for example, color, texture, features and surface of an object. The application of this technology spread's from medical science to national security including a vast variety of fields. In this system the main challenge is image indexing and similarity ranking for the proper retrieval. A solution to this problem is introducing a Firefly optimization combined Decision Tree (FF-DT) classifier to reduce the computational complexity in the classification stage of feature extraction. Precision rate and recall rate are the evaluation matrices used to evaluate the efficiency of this work. The result obtained from the experiment proves that the proposed approach is more efficient in CBIR system than the existing methods.