摘要
单一生物特征在识别过程中具有一定的局限性,基于人脸和人耳在位置上具有一定的关联性,提出了人脸人耳特征级融合的识别算法。采用主成分分析法(PCA)对人脸及人耳进行特征提取,然后运用稀疏表示对所提取的特征进行分类表达。而基于稀疏表示的人脸人耳识别方法,在遮掩、含噪声的情况下取得了不错的识别效果。实验证明,基于稀疏表示的人脸人耳融合的识别算法,具有较好的识别准确度。
A single biometrics has certain limitations in the identification process.This paper proposes a biometric fusion algorithm for human ear based on the correlation between human ear and human ear.In this paper,principal component analysis(PCA)is used to extract features of human face and human ear,and then sparse representation is used to classify the extracted features.The sparse representation based face recognition method has achieved good results in the case of occlusion and noise.Experimental results show that the recognition algorithm based on sparse representation of human ear fusion has better recognition accuracy.
作者
郑秋梅
马茂东
王风华
孙燕翔
李波
ZHENG Qiumei;MA Maodong;WANG Fenghua;SUN Yanxiang;LI Bo(College of Computer & Communication Engineering,China University of Petroleum, Qingdao 266580)
出处
《计算机与数字工程》
2019年第7期1640-1643,1661,共5页
Computer & Digital Engineering
基金
国家自然科学基金项目“稀疏表示结合质量评价的多模生物特征识别研究”(编号:61305008)资助
关键词
稀疏表示
主成分分析
多生物识别
特征融合
sparse representation
principal component analysis
multiple biometrics
feature fusion