摘要
为了解决当前人脸识别考勤系统在面对表情、眼镜、头发干扰时,其识别稳定性差的不足,论文设计了基于局部二值模式分类器(LBP Classifier)与特征脸(Eigen Faces)的人脸识别算法。首先,通过优化高维特征,提取低维特征,设计了基于LBP Classifier的人脸检测算子,标识出人脸区域;然后提取特征数据,结合人机交互输入,进行监督式学习,设计了基于EigenFaces的人脸识别算子,完成对人脸特征的识别。实验数据显示:与当前人脸识别算法相比,在面对表情、眼镜、头发干扰较大时,论文算法拥有更高稳定性与识别率。
In order to solve poor stability of its recognition of the current face recognition system of check on work attendance in the face of facial expression, glasses, hair interference, this paper designed face recognition system of check on work attendance based on LBP Classifier (local binary pattern Classifier with the EigenFaces (face). First of all, through the optimization of high-dimensional feature, low dimensional characteristics is refined, the face detection is designed based on LBP Classifier operator, face region is identified. And then the feature data is extracted, humanmachine interactive input is combined for supervised learning, design the face recognition is decigned based on EigenFaces operator. Finally, the recognition of face charcteristics is finished. Experimental data shows that compared with the current face recognition algorithm, in the face of the large interference of the expressions, glasses, hair this algorithm has higher stability and recognition rate.
出处
《计算机与数字工程》
2016年第8期1576-1580,共5页
Computer & Digital Engineering
关键词
人脸识别
局部二值模式分类器
特征脸
低维特征
监督式学习
face recognition, local binary pattern classifier, feature face, low dimensional feature, supervised learning