摘要
针对自动指纹识别系统中,关于时间复杂度高的指纹图像预处理算法进行研究,为了提高识别速度,有效去除噪声,提出一种有效的快速指纹预处理算法。算法基于指纹的方向场,采用查表的方法,对指纹图像进行8方向滤波增强,并采用混合式快速细化算法完成指纹的细化。对100幅URU4000B指纹采集仪采集的指纹图像在Matlab中进行仿真,结果表明92%的图像都能达到较好的处理效果。算法速度快,适用于配置一般的基于嵌入式平台的指纹自动识别系统。
This paper proposes an effective fast fingerprint pre-processing algorithm based on the thorough study of the fingerprint pre-processing algorithm which has the highest time complexity in AFIS.The algorithm adopts the method of look-up to enhance the fingerprint in 8 directions,and employs mixed fast thinning algorithm to thin the fingerprint.The 100 fingerprint images acquired by URU4000B are simulated in Matlab environment,and the results show that 92% of the images can achieve better effect.The pre-processing algorithm is very fast and can be used in AFIS based on normal embedded systems.
出处
《计算机仿真》
CSCD
北大核心
2010年第8期255-258,341,共5页
Computer Simulation
关键词
指纹增强
滤波
二值化
细化
方向图
Fingerprint enhancement
Filter
Binarization
Thinning
Directional image
