基于模板的指纹图像细化算法  

Fingerprint image thinning algorithm based on template

作　　者：蔡秀梅[1] 孙鹏[1] 

机构地区：[1]西安邮电大学自动化学院,陕西西安710121

出　　处：《西安邮电大学学报》2016年第3期59-63,共5页Journal of Xi’an University of Posts and Telecommunications

基　　金：国家自然科学基金资助项目(61571361);西安市科技计划资助项目(CXY1516(1))

摘　　要：为了提高指纹图像细化处理后的效果,给出一种基于模板的指纹图像细化算法。在逐像素细化(One Pass Thinning Algorithm,OPTA)算法的基础上,构造9个消除模板和7个保留模板;通过迭代方法遍历图像中所有像素,当遇到目标点时,提取该点周围的15-邻域像素;将提取出的像素分别与两套模板匹配,以判断该目标点能否被删除。所给算法在不做任何额外前、后处理的前提下,能够对纹线完全细化,处理后图像光滑无毛刺。In order to improve the effect of fingerprint image thinning process,a fingerprint image thinning algorithm based on template is proposed.Based on the one pass thinning algorithm（OPTA）,nine elimination template and seven reserved template are constructed,then,all pixels in the image are scanned by an iterative method to find out the target point,and the fifteen pixels around it are picked up.Match these pixels with the elimination templates and reserved templates respectively to determine whether the target point can be deleted or not.Without any additional pre-processing and post-processing,the proposed algorithm can fully refined the ridge line,and the processed image is smooth with little burr.

关 键 词：指纹图像 指纹细化 消除模板 保留模板 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]