基于ICA的眉毛识别方法研究  

Eyebrows Recognition Based on Independent Component Analysis

作　　者：李颜瑞[1] LI Yanrui(Information Engineering Department,Shan Xi Institute of Mechanical and Electrical Engineering,Changzhi,Shanxi 046011,China)

机构地区：[1]山西机电职业技术学院信息工程系,山西长治046011

出　　处：《信息记录材料》2021年第2期196-198,共3页Information Recording Materials

摘　　要：本文提出了一种基于ICA的眉毛识别方法。在经过预处理的眉毛图像上,先用ICA算法对眉毛图像进行特征提取,得到其特征向量,然后用K-近邻算法进行识别,最后经过实验验证,眉毛识别的识别率可以达到94.7%。再一次验证了,眉毛用于生物特征识别的可行性。This paper presents an eyebrow recognition method based on Partitioning Independent Component Analysis Recognition.In the pre-processed eyebrow image,Independent Component Analysis Algorithm is used to extract the feature vector of the eyebrow image,and then k-nearest neighbor algorithm is used to recognize the feature vector,Through experimental verification,the recognition rate of eyebrows can reach 94.7%.Once again,eyebrows are useful for Biometric identification.

关 键 词：独立分量分析 眉毛识别 K-近邻法则 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]