Summary
Palm lines are one of the most important features of palmprints. In this chapter we have proposed a novel approach on palm line extraction and matching for personal authentication. A set of directional line detectors has been devised for effective palm line extraction. To preserve the details of the lines’ structure, palm lines are represented by their chain code and then palmprints are matched by matching the points on their palm lines. The experiment results from a general database (DB1) demonstrate the proposed approach is more powerful for palmprint verification than the 2-D Gabor algorithm in [65] when FRR <0.91%, and the EER of our approach is also decreased from 0.6% of 2-D Gabor algorithm to 0.4%. Another experiment for testing its robustness against dirty palms (DB3) confirms the advantage of our approach over 2-D Gabor algorithm, where the EER of our approach is 0.63% less than that of 2-D Gabor algorithm. In addition, the average memory requirement for a palmprint is only 267 bytes and the average processing time, including preprocessing of palmprint image and matching, is only 0.6 second, which has proved the practical value of our approach. In conclusion, the proposed approach can effectively identify a person based on his/her palm lines which can be used on a real world application.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
Rights and permissions
Copyright information
© 2004 Springer Science + Business Media, Inc.
About this chapter
Cite this chapter
(2004). Line Features Extraction and Representation. In: Palmprint Authentication. International Series on Biometrics, vol 3. Springer, Boston, MA. https://doi.org/10.1007/1-4020-8097-2_8
Download citation
DOI: https://doi.org/10.1007/1-4020-8097-2_8
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4020-8096-8
Online ISBN: 978-1-4020-8097-5
eBook Packages: Springer Book Archive