摘要
提出了一种基于支持向量机体温检测模型的活性判别算法。该算法通过大量的人体前额和腋下温度的样本来训练基于支持向量机的体温检测模型,然后在活性判别过程中,通过检测被识别人的体温来完成活性判别的过程。由于人脸照片的温度不可能与正常人的体温相同,所以照片就被成功的排除在了人脸识别系统的外面,从而提高了人脸自动识别系统的安全性,而且在活性判别过程中,被识别人不需要做表情或者是姿态的变化来配合识别,极大的方便了被识别者,增强了人脸识别系统的方便性。
A liveness check algorithm based on body temperature measurement model using SVM (support vector machine) is proposed. The body temperature measurement model is trained by a large number of examples with forehead and oxter temperatures. And in the liveness check procedure, the users' body temperature is measured to get through the liveness check. Because the temperature of face photos can not be the same as the body temperature of the normal person, the photos are successfully stopped before entering the face recognition stage. That raised the securities of face automatic recognition system. And in the liveness check procedure, the users do not need to do the change of the expression or the posture cooperates with recognition. The users use this system conveniently, and the conveniences of face recognition systems is strengthened.
出处
《计算机工程与设计》
CSCD
北大核心
2007年第3期642-644,共3页
Computer Engineering and Design
关键词
人脸识别
活性判别
体温检测模型
支持向量机
样本
face recognition
liveness check
body temperature measurement model
support vector machine
example