An efficient palmprint identification system based on an indexing approach

Performance of a biometric based person identification system depends upon the search time and accuracy of identification. These two factors increase with the increase in the size of the database. One way to reduce the search time is to retrieve small number of candidate identities from the database against which matching is performed for person identification. In this paper we propose an approach for reducing the search space by indexing the database prior to identification process to improve the accuracy and search time. Experiments results on a palmprint database indicate 80% reduction in search space with 100% correct index power. Results reveal that the proposed method reduces the search time without compromising the accuracy of identification. The main objective is to look at suitability of an approach which searches and identifies a query image in reasonable time and with good accuracy.