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
- Forced Condensation of Cyanoacrylate With Temperature Control of the Evidence Surface To Modify Polymer Formation and Improve Fingerprint Visualization
- A Case Study Evaluating the Efficacy of an Ad Hoc Hospital Collection Device for Fentanyl in Infant Oral Fluid
- Observing Variation of Acoustical Characteristics of Several Common Firearms in a Quasi Anechoic Environment at a High Sampling Rate