首页 | 本学科首页   官方微博 | 高级检索  
     

基于L0稀疏约束的近似稀疏解人耳识别*
引用本文:田 莹,李雪滢,张德斌.基于L0稀疏约束的近似稀疏解人耳识别*[J].计算机应用研究,2017,34(6).
作者姓名:田 莹  李雪滢  张德斌
作者单位:辽宁科技大学 软件学院,辽宁科技大学 软件学院,辽宁科技大学 软件学院
基金项目:辽宁省教育厅资助项目:基于2D合成图像的多姿态人耳识别(L2014115)
摘    要:传统的人耳识别算法在人耳图像遮挡、噪声和人耳多姿态变化中表现出低识别率,近年来稀疏表示在模式识别领域中取得很好的成果。决定稀疏分类器识别精确度的因素主要是稀疏解的稀疏度。而稀疏度的估计就是稀疏向量中非0元素的估计,即向量L0范数。因此在人耳稀疏分类算法的研究中引入L0范数稀疏约束。综上所述,采取基于SRC(Sparse Representation-based Classification)稀疏模型,选取对人耳姿态变化具有强鲁棒性的特征逼近过完备字典,然后使用OMP(Orthogonal Matching Pursuit)算法直接解L0问题,并加入稀疏约束,从优化稀疏解的角度对人耳稀疏分类算法进行改进,提高人耳识别效率。

关 键 词:SRC稀疏分类  OMP算法  L0稀疏约束
收稿时间:2016/3/18 0:00:00
修稿时间:2016/5/24 0:00:00

Human ear recognition approximating sparse solution based on L0 sparse constraint
Tian Ying,Li Xueying and Zhang Debin.Human ear recognition approximating sparse solution based on L0 sparse constraint[J].Application Research of Computers,2017,34(6).
Authors:Tian Ying  Li Xueying and Zhang Debin
Affiliation:University of Science and Technology Liaoning,University of Science and Technology Liaoning,University of Science and Technology Liaoning
Abstract:The traditional human ear recognition algorithm showed low recognition rate on ear image block, noise and multi-profile ear. In recent years, sparse representation have made great achievements in the field of pattern recognition. However, sparse degree of solution is the main factors to decide Sparse classifier recognition accuracy. Sparse estimation was the estimation of the Non-zero elements in the sparse vector, namely Vector L0 norm. Therefore, L0 norm sparse constraint is introduced into the study of human ear sparse classification algorithm. In summary, this paper took the characteristic of the human ear which has a strong robust posture change of approach over complete dictionary, then taking solution using OMP algorithm for L0 questions and join the sparse constraint from the perspective of the optimized sparse to improve the human ear sparse classification algorithm. Improve the ear recognition efficiency.
Keywords:SRC sparse classification  OMP algorithm  L0 sparse constrain
点击此处可从《计算机应用研究》浏览原始摘要信息
点击此处可从《计算机应用研究》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号