基于2DLDA与SVM的人耳识别算法

Ear recognition algorithm based on 2DLDA and SVM

文中提出将二维线性鉴别分析(2DLDA)和支持向量机(SVM)相结合的人耳识别算法。先利用二维线性鉴别分析提取人耳图像的特征,再采用一对一的方法用支持向量机分类器实现人耳特征的分类识别。与传统的2DLDA对比实验结果表明,该方法具有更高的识别率,是一种非常有效的识别方法。

In this paper, we propose a human ear recognition algorithm which combined two-dimensional linear discrimination analysis（2DLDA） and support vector maehine（SVM）. First of all, use 2DLDA for feature extraction in the human ear image, then applying one-against-one method, use SVM classifier for the classification recognition according to the human ear feature. Compared with classical 2DLDA, experimental results show that our method has higher recognition rate and it is a very efficient recognition method.