[1]刘裕丰,许成哲*.基于TV-L2模型的人脸识别光照正规化算法[J].延边大学学报(自然科学版),2016,42(01):60-64.
　LIU Yufeng,XU Chengzhe*.Research of illumination normalization algorithm for face recognition based on TV-L2 model[J].Journal of Yanbian University,2016,42(01):60-64.
点击复制

基于TV-L2模型的人脸识别光照正规化算法

《延边大学学报(自然科学版)》[ISSN:1004-4353/CN:22-1191/N] 卷: 第42卷 期数: 2016年01期 页码: 60-64 栏目: 应用科学研究 出版日期: 2016-03-20
Title:
Research of illumination normalization algorithm for face recognition based on TV-L2 model
作者:
刘裕丰;  许成哲*
延边大学工学院 电子信息通信学科, 吉林 延吉 133002
Author(s):
LIU Yufeng;  XU Chengzhe*
Department of Eectronic and Communication Engineering, College of Engineering, Yanbian University, Yanji 133002, China
关键词:
人脸识别;  TV-L2模型;  纹理信息;  可变光照
Keywords:
face recognition;  TV-L2 model;  texture information;  varying lighting
分类号:
TP391.41
文献标志码:
A
摘要:
提出一种基于TV-L2模型的光照正规化算法,该方法提取侧重于保留人脸纹理特征信息的光照反射成分.首先利用TV-L2模型对人脸图像执行滤波,估计出光照成分后,根据反射表征模型在对数域中求取人脸反射成分图像,最后对人脸反射成分图像进行标准化之后得到独立于光照变化的图像.实验结果表明,本文提出的方法运行速度快,能有效地消除光照变化的影响.
Abstract:
This paper proposes an illumination normalization algorithm based on TV-L2 model. The presented method is utilized to extract the illumination reflectance components which focus on retaining the facial texture feature information. First, we perform the filtering on face images using TV-L2 model to estimate the illumination components. Then, obtain the reflectance components of face images in logarithmic domain based on reflection model. Finally, illumination-independent images are made by normalizing reflectance components of face images. The experimental result shows that the proposed method requires less computational time and can efficiently eliminate the influence of illumination variation.

参考文献/References:

[1] Zou X, Kittler J, Messer K. Illumination invariant face recognition: a survey[C]//First International Conference on BTAS. IEEE, 2007:1-8.
[2] Makwana R M. Illumination invariant face recognition: a survey of passive methods[J]. Procedia Computer Science, 2010,2:101-110.
[3] Chen T, Yin W, Zhou X S, et al. Total variation models for variable lighting face recognition[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2006,28(9):1519-1524.
[4] Park Y K, Park S L, Kim J K. Retinex method based on adaptive smoothing for illumination invariant face recognition[J]. Signal Processing, 2008,88(8):1929-1945.
[5] Xu C, Kim I. Simplified adaptive smoothing filter for face recognition[C]//Fourth International Joint Conference on Computational Sciences and Optimization. IEEE, 2011:341-344.
[6] Vishwakarma V P. Illumination normalization using fuzzy filter in DCT domain for face recognition[J]. International Journal of Machine Learning and Cybernetics, 2013,6(1):17-34.
[7] Xu C. Illumination invariant face recognition using nonlocal total variation in logarithmic domain[C]//International Conference on Measurement, Instrumentation and Automation. TTP, 2012:1652-1658.
[8] Rudin L I, Osher S, Fatemi E. Nonlinear total variation based noise removal algorithms[J]. Physica D: Nonlinear Phenomena, 1992,60(1):259-268.
[9] Georghiades A S, Belhumeur P N, Kriegman D J. From few to many: illumination cone models for face recognition under variable lighting and pose[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2001,23(6):643-660.

备注/Memo

收稿日期: 2016-01-24 基金项目: 延边大学青年基金资助项目(延大科合字[2015]第14号) *通信作者: 许成哲(1975—),男,博士,副教授,研究方向为信号处理、图像处理及模式识别.

更新日期/Last Update: 2016-01-20