Security Investigate

Iris Feature Extraction for Personal Identification Using Lifting Wavelet Transform

Download now Free registration required

Executive Summary

With an increasing emphasis on security, automated personal identification based on biometrics has been receiving extensive attention over the past decade. Iris recognition, as an emerging biometric recognition approach and has become a popular research and practical applications in recent years due to its reliability and nearly perfect recognition rates. In this paper, the authors present a novel approach to iris recognition. Their goal is to develop a Lifting (integer) wavelet-based algorithm that enhances iris images, reduces noise to the maximum extent possible, and extracts the important features from the image. Then the similarity between two iris images is estimated using Euclidean distance and comparison of threshold.

  • Format: PDF
  • Size: 363.86 KB