Efficient Classifying and Indexing for Large Iris Database Based on Enhanced Clustering Method

Explosive growth in the volume of stored biometric data has resulted in classification and indexing becoming important operations in image database systems. A new method is presented in this paper to extract the most relevant features of iris biometric images for indexing the iris database. Three transformation methods DCT, DWT and SVD were used to analyse the iris image and to extract its local features. The clustering method shouldering on the responsibility of determining the partitioning and classification efficiencies of the system has been improved. In the current work, the new Weighted K-means algorithm based on the Improved Firefly Algorithm (WKIFA) has been used to overcome the shortcomings in using the Fireflies Algorithm (FA). The proposed method can be used to perform global search and exhibits quick convergence rate while optimizing the initial clustering centers of the K-means algorithm. From the experimental results, the proposed method was indeed more effective for clustering and classification and outperformed the traditional k-mean algorithm. The Penetration Rates underwent reductions and reached the levels of 0.98, 0.13 and 0.12 for three different databases. Also, the Bin Miss Rates decreased to 0.3037, 0.4226 and 0.2019 for the investigated databases.