Title :
A Hidden Markov Model For Iris Recognition Method
Author :
Tong, Wang ; Pi-Lian, He
Author_Institution :
TianJin Univ., Tianjin
fDate :
May 30 2007-June 1 2007
Abstract :
Iris identification system is mainly composed of iris image acquisition, iris image preprocessing and iris image matching. Iris image matching is the key step to the system and effects on the precision and efficiency of the whole system directly. Iris image matching is mainly based on its texture, which can be present by the orientation field. An iris, which has the different orientation angle structure in different area and has a texture pattern correlation with the neighborhood areas, can be viewed as a Markov stochastic field. In this paper, a novel method based on Hidden Markov model that is used to model the orientation field of irises is present. The accurate and robust iris image matching can be achieved by matching Hidden Markov model parameters, which are produced and trained after the processing of extracting eigenirises to form the samples of observation vectors.
Keywords :
biometrics (access control); eigenvalues and eigenfunctions; eye; hidden Markov models; image recognition; image texture; stochastic processes; Markov stochastic field; eigenirises; hidden Markov model; image texture; iris identification system; iris image acquisition; iris image matching; iris image preprocessing; iris recognition method; orientation angle structure; texture pattern correlation; Biometrics; Educational institutions; Fingerprint recognition; Hidden Markov models; Image matching; Iris recognition; Pattern matching; Robustness; Surface texture; Waveguide discontinuities; Eigeniris; Hidden Markov Model; Iris Recognition; Orientation Field;
Conference_Titel :
Control and Automation, 2007. ICCA 2007. IEEE International Conference on
Conference_Location :
Guangzhou
Print_ISBN :
978-1-4244-0817-7
Electronic_ISBN :
978-1-4244-0818-4
DOI :
10.1109/ICCA.2007.4376669
