Face Recognition Based on Improved LTP

DOI

10.2991/ismems-17.2018.2How to use a DOI?

Keywords

Face recognition; LTP; ALTP; FERET database

Abstract

Localized binary model (LBP) is an efficient local feature description operator. As a nonparametric description operator, it has received more and more attention and has achieved great success in the field of face recognition. In this paper, we introduce only the first-order non-directional feature of LBP operator, and introduce the high-order differential ULDP operator in four directions, and apply the preprocessing method to face recognition. In addition, the threshold for the LBP operator is completely dependent on the defects of the central pixels. In this paper, the local threshold model (ALTP) of the adaptive threshold is proposed. The threshold of the region is automatically generated by calculating the mean and variance of the local region pixels. The experimental results in several commonly used face databases show that ULDP and ALTP in this paper have good robustness to face recognition in non-constrained environment, especially face recognition with illumination change.

Copyright

© 2018, 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  - Jiyuan Wang
AU  - Ruyang Zhang
AU  - Tingting Wu
AU  - Sooyol Ok
AU  - Eungjoo Lee
PY  - 2017/11
DA  - 2017/11
TI  - Face Recognition Based on Improved LTP
BT  - Proceedings of the International Symposium on Mechanical Engineering and Material Science (ISMEMS 2017)
PB  - Atlantis Press
SP  - 6
EP  - 10
SN  - 2352-5401
UR  - https://doi.org/10.2991/ismems-17.2018.2
DO  - 10.2991/ismems-17.2018.2
ID  - Wang2017/11
ER  -