Abstract
A biometric identification of persons wchich utilize contour of a human hand belogs to still very interesting and still not totally explored areas and its accuracy and effectiveness depends on technical capabilities to some extent. Presented paper solves given problem using combination of different algorithms. A hand contour is used, topological description of the hand, evolutionary algorithm, algorithm linear regression to estimate the knuckles positions and for contours comparison is used an algorithm Iterative Closest Point (ICP) in its genuine shape. All 5 fingers is at computer classification fully moveable, thumb has 2 knuckles. Modern evolutionary optimizers enable markedly to cut down computational demands of the algorithm ICP. Experimental verification of proposed recipes were performed with use of two different databases named THID and GPDS with persons of both gender and different age (cca 20-65let) with total number of oeprons in individual database 104 and 94. Experimental results checked succesfuly suitability of use combination of methods ICP and evolutionary optimizer which is named as EPSDE for solving of the given task with algorithmic complexity O(N) and success rate give by coefficient THID:EER=0.38% and GPDS:EER=0.35% on real images.
Highlights
The Persons identification problem with use of hand contour [8, 7, 14, 29, 33, 50] is very specific
The problem of hand contour classification is defined as the capability of classifier to find, as near as possible, the correspondence between template hand contour, and contour which is submitted for identification
All five fingers of the template contour of a hand are moveable, and thanks to that the classification process is more difficult compared with a hand contour which has only four fingers
Summary
The Persons identification problem with use of hand contour [8, 7, 14, 29, 33, 50] is very specific. A method described in the presented paper modifies, in a suitable way, a commonly used scheme of Euclidean metrics It uses a modern, and many years proved, evolutionary optimizer of 3rd generation and an Iterative Closest Point (ICP) algorithm [10]. Many years proved, evolutionary optimizer of 3rd generation and an Iterative Closest Point (ICP) algorithm [10] Thanks to this combination the method is capable of coming up with results of methods which are based on the Linear Discriminant Analysis (LDA) algorithm, which is at this time assumed as one of the best possible solutions for the given task and which is used in many papers e.g. in [8, 35]. For experimental purposes two different databases were used: Technocampus hand Image Database (THID) [20, 23, 58] and database GPDS [22, 55, 59] The Grupo de Procesado Digital de la señal, GPDS (Digital Signal Processing Group) with DPDS (División de Procesado Digital de la Señal) from Instituto para el Desarrollo Tecnológico y la Innovación en Comunicaciones IDeTIC, University of Las Palmas de Gran Canaria, Spain
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
