Performance Analysis of MSB Based Iris Recognition Using Hybrid Features Extraction Technique
Sunil Swamilingappa Harakannanavar1, Prashanth C R2, Raja K B3

1Sunil Swamilingappa Harakannanavar, Department of Electronics and Communication Engineering, S. G. Balekundri Institute of Technology, Belagavi, Karnataka, India.
2Prashanth C R, Department of Telecommunication Engineering, Dr. Ambedkar Institute of Technology, Bangalore, Karnataka, India.
3Raja K B, Department of Electronics and Communication Engineering, University Visvesvaraya College of Engineering, Bangalore, Karnataka, India.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP: 230-239 | Volume-8 Issue-6, August 2019. | Retrieval Number: E7292068519/2019©BEIESP | DOI: 10.35940/ijeat.E7292.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: In the modern days, biometric identification is more promising and reliable to verify the human identity. Biometric refers to a science for analyzing the human characteristics such as physiological or behavioral patterns. Iris is a physiological trait, which is unique among all the biometric traits to recognize an individual effectively. In this paper, MSB based iris recognition based on Discrete Wavelet Transform, Independent Component Analysis and Binariezed Statistical Image Features is proposed. The left and right region is extracted from eye images using morphological operations. Binary split is performed to divide the eight-bit binary of every pixel into four bit Least Significant Bits and four bit Most Significant Bits. DWT is applied on four bit MSB to extract the iris features. Then ICA is applied on approximate sub band to extract the significant details of iris. The obtained features are then applied on BSIF to obtain the enhanced response with final features. Finally features produced are matched with the test features using Euclidean distance classifier on CASIA database. The experiments are performed on proposed iris model using MATLAB 7.0 software considering various combinations of Person inside Database (PID’s) and Person outside Database (POD’s) to evaluate the recognition accuracy of the proposed iris model.
Keywords: Biometrics, Discrete Wavelet Transform, Independent Component Analysis, Binariezed Statistical Image Features, Euclidean Distance.