摘要
生物特征识别技术是通过人体本身所固有的、唯一的生理特征来自动地识别一个人。针对掌纹图像具有终身不变性、唯一性、主要特征明显、细节特征唯一等特点,论文提出了一种有效提高掌纹识别率的方法,即结合图像中的主元特征、不变矩特征和线特征的提取方法。该方法与单独使用主元特征提取、不变矩特征提取和线特征提取的方法进行了比较实验,结果证明新方法的识别率有明显提高。
Biometric identification technology is to automatically identify a person by inherent and only physiological characteristics in the human body itself. In the light that the palmprint image is life longly unchangeable, clear in main features, and unique in particulars, this paper proposes an effective methool for raising palmprint identification, which combines extractions of PCA features, invariable moment features and line features thereby enhancing Palmprint Identification feature matching rate. As compared to the method which adopts PCA features extraction, invariable moment features extraction, or line features extraction alone, the proposed method has markedly improved the identification rate.
出处
《信息安全与通信保密》
2007年第11期46-48,共3页
Information Security and Communications Privacy
关键词
掌纹识别
主元特征
不变矩
线特征
palmprint identification
PCA features
invariable moment features
line features
