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
- Analyzing and interpreting deoxyribonucleic acid from multiple donors using a forensically relevant single-cell strategy
- Just Science Podcast: Just Using Inadvertently Photographed Ridge Detail as Evidence
- Atmospheric Chemistry of Chloroprene Initiated by OH Radicals: Combined Ab Initio/DFT Calculations and Kinetics Analysis