摘要
具有生物背景的脉冲耦合神经网络具有自适应提取指纹特征的特性,基于此,首次提出了一种指纹图像特征提取的新方法——自适应耦合神经网络点火统计图的,此图不仅包含了指纹图像的灰度特征,还包含相邻像素之间的几何位置信息。此方法具有运算速度快及对旋转、平移、尺度不变性,是许多指纹特征提取算法不具备的优点。最后给出了部分实验的结果,以验证该方法的有效性.
Pulse coupled neural network (PCNN) which has biology backgrand is characterized with adaptive image separation and image feature extraction. In view of this thought, a new method of feature extraction of fingerprint named PCNN adaptive firing series statistical image is proposed in th is paper for the first time, it is pointed out that this image contains not only information about the gray feature of fingerprint, but also geometry information between adjacent pixels, which is more important and just the individual characteristics of the fingerprint which is type texture image.This method consists of segment and feature extraction has advantages of operate fast and invariability in the face of rotation, level move which many other methods of feature extraction of fingerprint have not. Finally the the effectiveness of the method is well verified by a large number of experiments showed some in this paper.
出处
《自动化与仪器仪表》
2011年第5期134-135,137,共3页
Automation & Instrumentation
关键词
脉冲耦合神经网络
神经元点火统计图
耦合特性
指纹特征提取
Pulse coupled neural network (PCNN)
PCNN adaptive firing series statistical image
Coupling
Fingerprint feature