A Frame Work for Reducing the Time for Image Retrieval with Genetic Algorithm

Abstract

In recent years especially in the last decade, the rapid development in computers, storage media and digital image capturing devices enable to collect a large number of digital information and store them in computer readable formats. The large numbers of images has posed increasing challenges to computer systems to store and manage data effectively and effi ciently. Although this area has been explored for decade sand many researches have been done to develop some algorithms that solve some of its problems, no technique has achieved the accuracy of human visual perception indistinguishing images. Nowadays, virtually all domains of human life including commerce, government, academics, hospitals, crime prevention, surveillance, engineering and historical research use information as images, so the volume of digital data is increasing rapidly. These images and their data are categorized and stored on computers and the problem appears when retrieving these images from storage media. Thus CBIR from large resources has become an area of wide interest in recent years especially in the last decade. To retrieve any image, we have to search for it among the database using some search engine. Then, this search engine will retrieve many of images related to the searched one. The main problem fo r t he user is the difficulty of locating his relevant image in this large and varied collection of resulted images. To solve this problem, text-based and content-based are the two techniques adopted for search and retrieval.The main objective of this paper is to build more generalized CBIR system which increase the searching ability and provide more accurate results. To improve the retrieval accuracy the system has taken the feedback from the user automatically. Here we used WANG database to evaluate the performance of the new system by calculating the precision and recall metrics. We also compared the new system with other existing CBIR systems. The performance of the new architecture in terms of average precision, recall and retrieval time has been shown to perform good. From he experimental results, it is evident that the new system has beaten other existing systems in terms retrieval time.