Classification of Gait Biometric on Identical Twins


Wan-Noorshahida Mohd-Isa, and Junaidi Abdullah, and Eswaran, Chikkanan (2012) Classification of Gait Biometric on Identical Twins. Journal of Advanced Computer Science and Technology Research, 2 (4). pp. 166-175. ISSN 2231-8275

[img] Text
Classification of Gait Biometric on Identical Twins.pdf
Restricted to Repository staff only

Download (863kB)


This is the first classification analysis of gait as a biometric on identical twins. In this paper, a small sample viability analysis is performed on using gait as a biometric in recognition of identical twins. The motivation behind this paper is that identical twins have high face similarities, where video-based surveillance system relying on face biometric alone may have difficulty in distinguishing between them. Our gait features are the angular displacement walking trajectories from gait videos. Next, this paper proposes to apply a trajectory normalization by a Bezier spline root-finding and re-sampling to deal with the unequal speed of the walking trajectories. Then, classification using linear discriminant analysis (LDA) and k-nearest neighbour are applied where the best average correct classification rate (CCR) is 76% when classifying an individual twin as a unique individual.

Item Type: Article
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75.5-76.95 Electronic computers. Computer science
Divisions: Faculty of Computing and Informatics (FCI)
Depositing User: Ms Suzilawati Abu Samah
Date Deposited: 01 Nov 2013 05:52
Last Modified: 01 Nov 2013 05:52


Downloads per month over past year

View ItemEdit (login required)