摘要
针对现行的一些细化方法在细化二值指纹图像时容易产生断裂和毛刺的状况,提出了一种基于正方形模板和三角形模板的脉冲耦合神经网络细化方法,并加入一种方向约束条件来避免经常出现的指纹脊线毛刺。实验结果表明,该算法不但能有效地缩短细化时间,而且细化后的指纹图像不会产生断裂和毛刺现象。此外,将该算法应用到自动指纹识别系统中后,不仅缩短了识别时间,而且识别率也得到了明显提高。
To improve the traditional fingerprint thinning methods,which prone to generate rupture and burr,a Pulse-Coupled Neural Networks（PCNNs） method based on square and triangle template is proposed for binary fingerprint image thinning.In addition,a direction-constraining scheme for avoiding fingerprint ridge spikes has been discussed.The simulation results show that the proposed method effectively shortens the thinning time and avoids the phenomenon of rupture and burr after thinning.Furthermore,after applying the algorithm to automated fingerprint identification system,the method not only saves the time of identifying,but also improves the recognition performance.
出处
《江南大学学报（自然科学版）》
CAS
2010年第2期138-142,共5页
Joural of Jiangnan University (Natural Science Edition)
基金
国家自然科学基金项目(60574051)
关键词
脉冲耦合神经网络
模板
方向约束条件
细化
pulse-coupled neural networks
template
direction constraints
thinning
