A Novel Approach for Generation of Optimized Image Index for Very Large Database

Abstract

The rapid and continuous growth in the multimedia technology, computational power and the availability of storage devices like NAS, the size of image databases increases day by day over the Internet through many academic, commercial and social websites. So, these image databases attracted the attention of the researchers for fast and efficient image retrieval. The mechanisms like indexing and hash table can be used to solve these problems. The main challenge is to find a proper index technique which improves query performance. This paper focuses on the novel approach for design and development of efficient indexing technique based on B+ tree. The privacy has been maintained by applying the AES encryption algorithm for the nodes. Genetic algorithm has been applied to generate the optimized B+ tree. The result analysis shows that the optimized B+ tree generated by our novel approach gives better result than the B+ tree based on different analysis parameters like number of nodes, memory, search time, insertion time etc.