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
- Superhydrophobic Surface Modification of Polymer Microneedles Enables Fabrication of Multimodal Surface-Enhanced Raman Spectroscopy and Mass Spectrometry Substrates for Synthetic Drug Detection in Blood Plasma
- Differential DNA Preservation of Thermally Altered Tissue and Bone
- Should Survey Likert Scales Include Neutral Response Categories?