In this paper, the authors propose a real-time video-based iris image processing method to eliminate the bad-quality video image frames, taking advantage of the correlations among video frames.
Iris recognition is an important method for identifying a person. Currently, most iris recognition methods are based on individual images. For non-cooperative user identification, video image-based methods can provide more information; however, the iris image quality may vary from frame to frame. (Publisher abstract provided)
Downloads
Similar Publications
- A Computational Study on the Atmospheric Fate of Carbon-Centered Radicals from the 3-Methyl-2-butene-1-thiol + •OH Reaction: Mechanistic Insights and Atmospheric ImplicationsArticle link copied!
- Latent Fingerprint Matching: Fusion of Rolled and Plain Fingerprints
- An Investigation of Intrinsic and Extrinsic Factors that Influence Soil Microbial Succession During Human Decomposition