Optimal feature set selection in online signature verification Online publication date: Thu, 30-Nov-2017
by Sudhir Rohilla; Anuj Sharma; R.K. Singla
International Journal of Biometrics (IJBM), Vol. 9, No. 4, 2017
Abstract: The online signature verification has attracted many researchers in recent past as it offers useful real life applications. This paper presents role of four types of feature sets as static, kinematics, structural and statistical in nature and these feature sets are analysed in context of online signature verification. The signatures are verified as single trajectory and in combination of multiple sub-trajectories. We have applied feature sets with all possible permutations to signature trajectory and sub-trajectories. We have computed a total of 80 features and categorised to four feature sets on the basis of their behavioural characteristics. The inter-valued symbolic representation technique has been used to clearly understand the impact of each individual feature set or in combinations of feature set. The simulation results are presented using popular benchmark dataset SVC 2004 where both sub-datasets as TASK1 and TASK2 are used. The experimental results show that it is a promising correlation between different feature sets and suggest the optimal combination among several combinations of feature sets.
Online publication date: Thu, 30-Nov-2017
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Biometrics (IJBM):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email firstname.lastname@example.org