摘要
提出使用特征脸和二叉树支持向量机(BT-SVM)分类器相结合的方法进行人脸识别。首先从训练图像中求得特征脸空间,然后将训练集和测试集图像投影到特征脸空间得到投影系数,使用训练样本投影系数训练BT-SVM分类器,再使用BT-SVM分类器对测试图像进行识别。在ORL人脸库进行模拟试验,结果表明BT-SVM分类算法获得比SVM分类算法更高的识别率。
The techniques of eigenfaces and Binary Tree Support Vector Machine（BT-SVM） classifier are combined to recognize face in this paper. A series of eigenfaces are firstly established from the training images, and the training and testing images are projected into the space spanned by the eigenfaces, producing the coefficients. The BT-SVM classifiers are built with training coefficients, which are＇used for classifying training and testing images, and recognition accuracy percentage values are calculated. The experiments are implemented with ORL face databases, and the results demonstrate that BT-SVM classifier has better performance than SVM.
出处
《华东交通大学学报》
2009年第2期47-51,共5页
Journal of East China Jiaotong University
基金
江西省教育厅科技资助项目(GJJ08237
GJJ09211)
作者简介
胡林峰（1977-），男，江西南昌人，讲师，研究方向为模式识别高性能计算技术。
