摘要
针对基于模板脉冲耦合神经网络(PCNN)指纹图像细化算法细化时间长、纹线断裂、细化不彻底等问题,通过增加4个细化模板,重新构造方形模板及改变细化过程,提出一种基于改进PCNN的指纹图像细化算法。实验结果表明,该算法能够较好地满足细化要求,细化彻底、速度快且纹线光滑无毛刺,能够应用于其他二值图像。
Aiming at the problem of slow thinning speed, ridge breaking, not thinning to one pixel in fingerprint image thinning algorithm using template-based Pulse Coupled Neural Network（PCNN）, four thinning templates are added, the rectangle templates are redesigned, the process is changed, and an improved thinning algorithm is presented. Experimental results show that this algorithm has many advantages such as complete thinning, high speed, smooth skeleton without spikes. It can be applied to other binary images.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第18期180-181,190,共3页
Computer Engineering
基金
江苏省高校自然科学研究基金资助项目(06KJD510167)
关键词
指纹图像
细化
脉冲耦合神经网络
模板
fingerprint image
thinning
Pulse Coupled Neural Network（PCNN）
template
