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 Effect of Ultraviolet Radiation on Dyed Man-Made Textile Fibers Using UV-Vis Microspectrophotometry (MSP): Technical Aspects on Spectral Alterations in Time
- Cross-validation and evaluation of the performance of methods for the elemental analysis of forensic glass by μ-XRF, ICP-MS, and LA-ICP-MS
- Human Decomposition Ecology at the University of Tennessee Anthropology Research Facility