摘要
针对掌纹识别中图像预处理所带来的位置变换,一种有效的模板学习算法被应用于掌纹识别的特征提取。通过样本学习,利用参数估计的方法,每种掌纹模式的理想模型表示为小波域内的特定参变量。模式识别的过程因此转换为模板之间的匹配。实验结果显示,该算法很好地解决了预处理算法的缺陷,并且获得了非常高的识别率。
An efficient template learning algorithm which aims at compensating the transformations as a result of palmprint preprocessing is used for feature extraction of palmprint. Each kind of palmprint pattern is described by parameters in wavelet domain that learned form samples by estimation. Therefore, the stage of pattern recognition is transmitted into the matching of templates. The experiment results suggest that this method solves the drawback of preprocessing in a large extent and achieves high recognition rate.
出处
《计算机应用研究》
CSCD
北大核心
2003年第7期64-66,共3页
Application Research of Computers
关键词
生物特征识别
小波变换
特征提取
模式识别
Biometrics
Wavelet Transform
Feature Extraction
Pattern Recognition
