基于奇异值特征向量的支持向量机人脸识别技术

A Face Recognition Method Using Support Vectors Machine Based on SVD Vectors

支持向量机的人脸识别的方法,是利用专门的小样本统计理论来解决典型小样本问题的人脸识别技术。该文作者采用对图像灰度变化有良好稳定性等特性的图像矩阵的奇异值(SVD)特征向量来表征人脸。实验取得了较高的识别率,不仅验证了方法的有效性,也说明了支持向量机在人脸识别领域的实用性。

The face recognition method using support vectors machine is a method for few-samples face recognition, which uses the few-samples statistic theory. It uses the SVD vectors of image matrixes with good characters as image gray changes to represent human faces, and it gets a satisfying result. The results of experiment show this method is effective and support vectors machine is practicable in human face recognition.