重庆理工大学学报（自然科学版）2013，Vol.27Issue(6)：84-87,4.DOI:10.3969/j.issn.1674-8425(z).2013.06.019
基于组合核函数支持向量机的人脸识别
Face Recognition Based on Combined-Kernel Function of Support Vector Machine
摘要
Abstract
Face image feature parameter was extracted by using K-L transform,then SVM was employed as face identification.The kernel function is the core of the SVM,and different kernel function will produce different classification effect.In order to get kernel function with good learning ability and generalization ability,and according to the properties,kernel function will be local kernel function and global kernel function is linear combination into a new kernel function-combined-kernel function.It used Cross-Validation method to optimize its parameters and combination coefficients.We applied the kernel in SVM and compared the common kernel in YALE face library and the ORL face database recognition experiment.The simulation results verify the effectiveness of the kernel function.Experimental results certify that the combined-kernel function of SVM is superior to common kernel function of SVM.关键词
支持向量机/人脸识别/组合核函数Key words
support vector machines/ face recognition/ combined-kernel function分类
信息技术与安全科学引用本文复制引用
晁学民,周继萍..基于组合核函数支持向量机的人脸识别[J].重庆理工大学学报（自然科学版）,2013,27(6):84-87,4.基金项目
重庆市自然科学基金资助项目(2010JJ0079,2012JJA0594,2012JJA0792) （2010JJ0079,2012JJA0594,2012JJA0792）
重庆市教委科技项目(KJ110614) （KJ110614）
