摘要
虹膜识别技术是当前生物特征识别领域的研究热点,不同的虹膜特征提取算法会直接影响虹膜识别性能。分别基于多通道Gabor滤波器算法、灰度共生矩阵(GLCM)算法和Haar小波变换、局部二值模式(LBP)算法对不同影响因素下采集到的虹膜图像进行特征提取并分类。实验结果表明,基于LBP特征和基于Haar小波变换特征的虹膜特征提取方法均具有较强的光照鲁棒性,且识别准确度很高,但对图像噪声的抗干扰能力很低。基于Gabor滤波器的虹膜识别相对更稳定,识别准确度与前两种虹膜识别技术很接近。基于GLCM的虹膜识别分类能力较差,但抗光照和图像噪声影响力较强。
Iris recognition technology is a research hotspot in the field of biometric recognition,the different iris feature extraction algorithms will directly affect the performance of iris recognition.The iris images collected under different influencing factors are extracted and classified based on multi-channel Gabor filter algorithm,grey level co-occurrence matrix(GLCM)algorithm,Haar wavelet transform and local binary patterns(LBP)algorithm,through comparative analysis,the results show that the iris feature extraction methods based on LBP feature and Haar wavelet transform feature have strong illumination robustness and high recognition accuracy,but the anti-interference ability to image noise is very low.The iris recognition based on Gabor filter is relatively stabler,and the recognition accuracy is very close to that of the first two iris recognition technologies.The iris recognition and classification ability based on GLCM is poor,but it has strong resistance to light and image noise.
作者
齐志坤
姜囡
徐浩森
QI Zhikun;JIANG Nan;XU Haosen(The College of Criminal Investigation Police University of China for Public Security Information Technology and Information,Shenyang,China)
出处
《光电技术应用》
2022年第3期48-57,80,共11页
Electro-Optic Technology Application
关键词
虹膜识别
特征提取
GABOR滤波器
GLCM
HAAR小波变换
LBP
iris recognition
feature extraction
Gabor filter
grey level co-occurrence matrix(GLCM)
Haar wavelet transform
local binary patterns(LBP)