摘要
提出了一种基于小波变换的眉毛识别方法.该方法利用小波变换进行眉毛特征提取,选取奇偶行三层小波变换的高频、水平分量、垂直分量和整体二层小波变换的低频部分作为特征,利用最近邻法则进行识别.实验结果对比表明,该方法简单且识别率较高.
A new method of eyebrow recognition based on wavelet transform is proposed. Eyebrow images are translated into feature extraction by wavelet transform, the third HH3 HL3 , LH3 components and the second LL2 are selected as eyebrows features, with the help of the nearest neighbor method for recognition. The experiment results show this method is simple and accurate.
出处
《云南民族大学学报（自然科学版）》
CAS
2014年第4期293-295,共3页
Journal of Yunnan Minzu University:Natural Sciences Edition
关键词
眉毛识别
小波变换
特征提取
最邻近距离法则
eyebrow recognition
wavelet transform
feature extraction
nearest neighbor method
