1 November 2003 Node-matching-based pattern recognition method for retinal blood vessel images
Author Affiliations +
Optical Engineering, 42(11), (2003). doi:10.1117/1.1613278
Abstract
Human retinal blood vessel images possess certain characteristics that are suitable for personal identification and other advanced security applications. We propose an image processing algorithm for matching retinal blood vessel patterns based on the locations of intersections of blood vessels. The algorithm includes preprocessing, orientation registration, and similarity measurement. Preliminary experimental results show that our method achieves good discrimination among images from different subjects.
Tusheng Lin, Yibin Zheng, "Node-matching-based pattern recognition method for retinal blood vessel images," Optical Engineering 42(11), (1 November 2003). https://doi.org/10.1117/1.1613278
JOURNAL ARTICLE
5 PAGES


SHARE
RELATED CONTENT

Longitudinal analysis of mouse SDOCT volumes
Proceedings of SPIE (March 13 2017)
Use of Gabor filters and deep networks in the segmentation...
Proceedings of SPIE (February 16 2017)
Registration Of Rotated Images
Proceedings of SPIE (June 03 1987)
New algorithm for matching 2D objects
Proceedings of SPIE (May 22 2002)
Object recognition in a digital image
Proceedings of SPIE (July 30 2002)
Hybrid Architecture For Real Time Image Registration
Proceedings of SPIE (January 18 1988)

Back to Top