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
- Computational Methods for the Interpretation of Forensic DNA Samples
- Targeted-Ion Mass Spectrometry for the Identification of Forensically Relevant Biological Fluids and Samples from Sexual Assault Evidence
- Determining the Precision of High-Throughput Sequencing and Its Influence on Aptamer Selection