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
- The Application of Proteomics to Forensic Human Identification and Disease Characterization Using Hair Shafts
- Just Science Podcast: Just Forensic Podiatry Supporting Investigations
- Resolution of Mitochondrial DNA Mixtures using a Probe Capture Next Generation Sequencing System and Custom Analysis Software (Dissertation)