Efficient Identification Based on Human Iris Patterns

This paper proposes a person identification system (PID), which works with non-illumination low-resolution eye images. Radial-Scan and threshold methods are used for Iris normalization and Iris segmentation from the eye image respectively. Conventional neural network is used for iris matching. The proposed algorithm works with 96 features of the iris. Space complexity is less and Iris codes are stored in an encrypted file. The Algorithm has been implemented and tested on 1500 iris patterns. Images include normal iris captured under in-door and out-door conditions and from disease affected eyes.

• Type: Conference
• Year: 2004
• Where: ICVGIP
• Authors: A. Chitra, R. Bremananth
• Keywords: Eye Image, ICVGIP 2004, ICVGIP 2008, Iris Normalization, Iris Patterns