融合指纹和人脸的生物特征身份认证方法

摘要

The main algorithms of two single-mode authentication system,based on face,fingerprint identification were introduced respectively.In fingerprint identification subsystem,the improved algorithm of OPTA(One-Pass Thinning Algorithm) was applied to thinning binary image of the fingerprint image.Then,the feature images were matched by the Hough transform.And in the face identification subsystem,a method combined the way of feature extraction based on 2D Gabor filter with the matching algorithm of 3 D face was applied.One two-mode biometrics fusion system was proposed.The linear fusion approach about the matching rate of each subsystem was applied in decision-making level.The experimental results demonstrate that the established system works safer with low False Acceptance Rate.%在指纹识别模块中采用改进的OPTA算法对指纹图像进行二值化和细化,采用基于Hough变换点模式匹配算法对特征图像进行匹配；在人脸识别模块中,采用基于二维Gabor滤波的特征提取法和三维人脸匹配方法相结合的识别方法.建立了融合指纹和人脸两种生物特征的识别系统,利用单一特征匹配的结果分数,在决策层采用线性方法进行融合.实验结果表明:建立的融合系统降低了错误接受率,提高了系统的安全级别.