3D palmprint data fast acquisition and recognition
Xiaoxu Wang, Shujun Huang, Nan Gao, Zonghua Zhang*
School of Mechanical Engineering, Hebei University of Technology, Tianjin, China. 300130
ABSTRACT
This paper presents a fast 3D (Three-Dimension) palmprint capturing system and develops an efficient 3D palmprint
feature extraction and recognition method. In order to fast acquire accurate 3D shape and texture of palmprint, a DLP
projector triggers a CCD camera to realize synchronization. By generating and projecting green fringe pattern images
onto the measured palm surface, 3D palmprint data are calculated from the fringe pattern images. The periodic feature
vector can be derived from the calculated 3D palmprint data, so undistorted 3D biometrics is obtained. Using the
obtained 3D palmprint data, feature matching test have been carried out by Gabor filter, competition rules and the mean
curvature. Experimental results on capturing 3D palmprint show that the proposed acquisition method can fast get 3D
shape information of palmprint. Some initial experiments on recognition show the proposed method is efficient by using
3D palmprint data.
Keywords: 3D biometrics; 3D palmprint; Gabor filter; Competition rules; Mean curvature.
1. INTRODUCTION
In recent years, with the rapid development of Internet and e-commerce, it is becoming an important and key issue how
to identify the user's identity to ensure the security of information system [1]. Due to easy access and stability, as a kind
of biometric feature, palmprint has been widely studied in academia and applied in many fields [2].
2D (Two-Dimension) palmprint identification technology has been applied to the handheld devices by Fujitsu.
Japanese Mitsubishi bank is combining palmprint recognition system and ATM machines in order to enhance security. In
many important departments such as hospital, police agency and airport, palmprint identification system has been
applied to strengthen the security. Palmprint recognition is still in an early period but has got rapid development in
China, for example, Dorcon palmprint recognition system, palmprint entrance guard control system at Changchun Long
Jia airport, and the first international high precise palmprint identification system developed at Hong Kong Polytechnic
University [3, 4].
Palmprint recognition is a burgeoning biometric identification technology. The various lines from the fingers to the
wrist part at the end of the palm are collectively referred to as palmprint. As an important biometrics, many study works
have been done to recognize human being. The existing researches and applications are based on the 2D palmprint
features. In fact, palmprint is 3D biological characteristics. Because of the lack of 3D information, 2D biometrics will
affect the precision of the subsequent authentication and identification. Because of the elastic deformation and the
varying value of the contact pressure by hand skin contact, the captured palmprints exist unpredictable nonlinear
deformation.
Many researchers improved the fault tolerance in the algorithm level to make up for the captured data quality.
However, the high quality biometric information acquisition is a crucial step for improving to overall performance of the
automatic identification system. The surface shape measurement of the 3D object is an important method of acquiring
the 3D biological characteristics with high accuracy. The texture of biological characteristic provides useful information
for certification and identify [5,6]. Due to the advantages of high resolution, fast acquisition speed, and automatic data
processing, the optical non-contact measurement is considered to be the most promising method to obtain 3D objects’
shape. Compared with 2D biological characteristic, the 3D palmprint has the following advantages: 1) Because of non-
contact measurement, there is no elastic deformation and nonlinear distortion; 2) It can obtain distribution pattern of
hand shape characteristic parameters without any distortion; 3) It can obtain the space coordinate geometry and
orientation; 4) The colourful texture on surface provides more information on the biological characteristics; 5)
Insensitive to the external environmental light.
Zhang et al. designed a 3D palmprint system by projecting phase-shifted structure light onto hand palm through a
DLP (Digital Light Processing) projector and capturing the palm image with structure light by a CCD cameras [7]. 3D
palmprint data were obtained by processing the captured images. The color texture information of the palm surface
Real-time Photonic Measurements, Data Management, and Processing, edited by Bahram Jalali, Ming Li, Keisuke Goda,
Mohammad Hossein Asghari, Proc. of SPIE Vol. 9279, 92790P · © 2014 SPIE ·
CCC code: 0277-786X/14/$18 · doi: 10.1117/12.2073028
Proc. of SPIE Vol. 9279 92790P-1
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 11/18/2014 Terms of Use: http://spiedl.org/terms