An Effective Content Based Image Retrieval Model using Improved Memetic Algorithm

Content Based Image Retrieval (CBIR) technique retrieves relevant images based on a query image (QI) from larger databases. The existing CBIR models filter only a minimal number of features and tends to decrease the efficiency. This paper proposes an effective CBIR model by the use of Improved Memetic Algorithm (IMA). At first, feature extraction process takes place when the features exist in the database are extracted and saved in a feature repository. These feature sets are comprised with color signatures along with the shape and color texture features. At the same time, the features present in the QI are also extracted. Then, new similarity estimation has been developed by applying a meta-heuristic technique named as IMA. The presented IMA-CBIR model is tested using a benchmark dataset and the results indicated that the proposed IMA-CBIRS are highly effective in terms of precision and recall when compared with existing techniques.