摘要
在兼顾实时性的情况下,为了进一步提高人脸识别的识别率,本文提出一种基于特征联合和支持向量机的人脸识别方法。首先,提取人脸样本的梯度方向直方图特征和局部二值模式特征,并将这两种特征进行联合形成样本的联合特征。其次,使用主成分分析法对样本联合特征进行降维得到样本的低维联合特征。最后,利用训练样本的低维联合特征训练支持向量机得到一个人脸识别器,并利用该人脸识别器对测试样本进行识别。基于ORL人脸库的实验结果表明,与现有方法相比,本文方法在取得较高识别率的同时也取得了较好的实时性。
In order to further improve the face recognition rate in the case of taking into acount of the real-time, a novel face recognition method based on features combination and support vector machine is proposed. Firstly, the sample features of histograms of oriented gradients and local binary patterns are extracted and combined as the sample＇s combined features. Secondly, the principal component analysis method is adopted to reduce the dimension of the sample＇s combined features and the low dimensional combin- ed features can be obtained. Finally, a support vector machine is trained by using the low dimensional combined features to form a face recognizer, and then the face recognizer is utilized to recognize the test samples. The experiments based on ORL face database show, compared with the existing methods, the proposed method can achieve better recognition rate and real-time.
出处
《燕山大学学报》
CAS
2012年第6期519-525,共7页
Journal of Yanshan University
基金
河北省自然科学基金资助项目(F2010001276)
关键词
人脸识别
梯度方向直方图
局部二值模式
支持向量机
ORL人脸库
face recognition
histograms oforiented gradients
local binary pattems
support vector machine
ORL face database
