摘要
为了提高指纹图像识别的准确性和可靠性,先对指纹图像采用主成分分析(PCA)算法进行预处理,去除指纹图像的二阶统计特征相关.用ICA算法对加噪的指纹图像进行提取识别,并利用统计方法确定混合矩阵,实验表明:在一定的条件下,ICA能够很好地分离混合图像,提高准确性和可靠性.
In order to improve the fingerprint image recognition accuracy and reliability. Firstly they deal with fingerprint image using principal component analysis （PCA） algorithm, which is proposed to remove the fingerprint image of second order statistic characteristics related preprocessing, Than by using statistical method in mixed matrix, we applied ICA algorithm to extract the adding noise fingerprint image, and the experimential result shows that ICA is able to separate mixed image accuracy and reliability.
出处
《数值计算与计算机应用》
CSCD
2013年第2期89-94,共6页
Journal on Numerical Methods and Computer Applications
关键词
图像处理
主成分分析
独立成分分析
指纹识别
image processing
principal component analysis
Independent componentanalysis
Fingerprint identification
