An Approach for Human Identification by Ear Biometric System
P. Swapnil1, Devesh Narayan2, Sipi Dubey3
1P. Swapnil, Computer Science and Engineering, Rungta College of Engineering and Technology, Bhilai, India.
2Devesh Narayan, Assoc. Prof., Computer Science and Engineering, Rungta College of Engineering and Technology, Bhilai, India.
3Dr. Sipi Dubey, Chairperson, Board of Studies, Computer Science and Engineering, CSVTU, Bhilai (C.G.), Professor in Department of CSE, at RCET, Bhilai, India.
Manuscript received on March 12, 2015. | Revised Manuscript received on March 13, 2015. | Manuscript published on March 25, 2015. | PP:14-18 | Volume-3 Issue-5, March 2015. | Retrieval Number: E0931033515
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Ear biometric for identification of human is quite complex task. It’s use either uni-modal or multi-modal approach in order to authenticate a person. A uni-modal biometric system involves a single source of biometric to identify a person. This paper is based on uni-modal approach with ear as a biometric trait for recognizing a person. In this paper, an experimental approach is used for identifying a person based by their ear structure. As researchers suggest that ear is a physiological biometric that is quite reliable source of identifying human, we here, present the work related to this field of ear biometrics. Firstly, we will discuss about identification process of our ear biometric system that includes pre-processing, edge detection using canny edge detection method and then extracting farthest boundary points that form a maxline. Now this maxline is divided into multiple points to extract geometrical intersecting points on the outer edge. Then finally template matching is done for identification.
Keywords: Biometrics, ear biometrics, canny edge detection, maxline and template matching.