A New Location Method of Touchless Knuckleprint based on Computer Vision

DOI

10.2991/icmeit-19.2019.103How to use a DOI?

Keywords

feature extraction; ROI segmentation; knuckleprint location.

Abstract

The knuckleprint positioning is the key basis for the biometric recognition. This paper introduces a new location method of touchless knuckleprint based on computer vision. Firstly, in order to reduce the light impact, the RGB image is converted into YCbCr color space. We use a hybrid algorithm of elliptical model and circular gradient to segment gesture area in YCbCr color space. Subsequently, an efficient algorithm for the extraction of fingertips and finger valleys is proposed based on the distance curve, which usually takes 1 to 3 filtering operations. After obtaining the fingertips and finger valleys, the region of interest (ROI) could be located with the quadrangular window shape. Finally, because the gradient value at the knuckleprint firstly decreases to a local minimum and then increases to a local maximum, the knuckleprint could be found from the gradient curve. The method was tested on a data set of 20 people (400 hand images), and the correct rate could reach 97 percent. Experiments show that the location method of touchless knuckleprint has good effects in terms of the time complexity, space complexity and accuracy.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Xinqi Wang
AU  - Yanping Hu
PY  - 2019/04
DA  - 2019/04
TI  - A New Location Method of Touchless Knuckleprint based on Computer Vision
BT  - Proceedings of the 3rd International Conference on Mechatronics Engineering and Information Technology (ICMEIT 2019)
PB  - Atlantis Press
SP  - 648
EP  - 654
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmeit-19.2019.103
DO  - 10.2991/icmeit-19.2019.103
ID  - Wang2019/04
ER  -