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
- A Quantitative Understanding of Uniqueness and Reproducibility of Firearm Toolmark Surfaces
- Frontal sinus morphology as a forensic identification method: a comparison of intra-observer scores between scout radiographs and 3D skull images
- Mass Spectrometric Insights into the Identification of Ignitable Liquids in Fire Investigations Using DART-MS and GC/MS Techniques