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
- Using Clonal Massively Parallel Sequencing to Characterize Heteroplasmy in the mtDNA of Human Head Hair, Pubic Hair, and Buccal Samples
- New Psychoactive Substances in Wastewater from Rest Areas and Commercial Weigh Stations along Interstate Highways
- Assessing the expanded capacity of modern μ-XRF SDD systems for forensic analysis through an interlaboratory study: Part II—Vehicle glass