Iris recognition on Hadoop: A biometrics system implementation on cloud computing

Abstract

Cloud computing is one of the highly researched areas today, with an objective of taking advantage of various computational resources. In this paper we have used cloud computing environment with the aim to speed up the matching process of biometric traits. We have used iris recognition, a biometric technique, as it is one of the strongest method of authentication. Also Iris recognition is stable over time. We have used Hadoop [1], an open source cloud computing environment, to develop this model. Hadoop implements Map/Reduce [1] framework in Java. Map/Reduce make easy to process large amount of data on cloud. The results shows that there is an effective speedup and efficiency gain of Iris template matching on Hadoop process over sequential process.