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
- Addressing the Challenges of Detecting Drug-Facilitated Sexual Assault
- Stage Transitions in Lucilia sericata and Phormia regina (Diptera: Calliphoridae) and Implications for Forensic Science
- Design of Light-Induced Solid-State Plasmonic Rulers via Tethering Photoswitchable Molecular Machines to Gold Nanostructures Displaying Angstrom Length Resolution