人脸图像的自动校准算法

An Automatic Normalization Algorithm for Face Images

人脸自动识别是计算机模式识别领域的一个活跃课题，有着十分广泛的应用前景。在人脸图像中人脸的尺度、位置与姿态变化影响了人脸特征的抽取，所以需要对人脸图像作人脸校准的预处理。文中提出的算法利用模板匹配方法从人脸图像中提取出人脸姿态校准图像，再从中抽取出两个眼睛的位置，从而对人脸的尺度、位置与姿态变化实现自动校准。对有２１６幅人脸图像的人脸图像数据库作人脸自动校准实验，成功率是１００％，而且利用所得到人脸脸部校准图像进行的人脸识别实验的正确识别率可达到９０％左右。

The automatic recognition of human faces is an active problem in the area of computer pattern recognition due to a wide range of potential applications. This paper presents an algorithm to get pose normalized image with template matching, to extract eye positions, and to realize automatically normalization of scale, position and pose changes for a face image. Automatic normalization experiments have been done with a human face image database of 216 images. The rate of normalizing successfully is 100%. Face recognition experiments with the norma lized face images have been also done. The recognition rate is about 90%. These results show that the algorithm is effective.