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
- Forensic Handwriting Examination and Human Factors: The Report of the Expert Working Group for Human Factors in Handwriting Examination
- Haplotype Data for 23 Y-Chromosome Markers in Four U.S. Population Groups
- Local Anesthetics (Lidocaine and other 'Caines): Toxic Adulterants Found in Illicit Street Drugs