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
- Analysis of DNA Typing Data for Forensic Applications
- Prediction of Skull Fracture Risk for Children 0-9 Months Old Through Validated Parametric Finite Element Model and Cadaver Test Reconstruction
- Evaluation of a 49 InDel Marker HID Panel in Two Specific Populations of South America and One Population of Northern Africa