摘要
指纹的自动分类对于提高检出速度和识别率有重要的意义。有效地利用指纹的纹理结构和纹理方向等固有信息,可以提高指纹的分类类别。基于二值指纹图像进行分类不完全依赖中心点定位的准确性,避免了利用三角点的干扰信息,应用脊线追踪算法进行指纹分类,赋予了中心脊线方向连续的变化,可以将指纹自动分类算法提高到34类。
Fingerprint automated classification largely contributes to improving identification rate and detecting speed, Making an effective use of intrinsic fingerprint features like texture structure and texture orientation can increase the total of categories being classified. Classifying based on binary fingerprint image incompletely depended on accuracy of core, avoiding the interferential information of using the delta. Applying the ridge following method to classify fingerprint could follow the change of fingerprint ridge orientation. Fingerprints could be classified into 34 categories.
出处
《计算机应用研究》
CSCD
北大核心
2008年第7期2190-2193,2196,共5页
Application Research of Computers
基金
国家自然科学基金资助项目(60675015)
国家科技型中小企业技术创新基金资助项目(05C26214101032)
关键词
指纹分类
二值图像
中心点
脊线追踪
fingerprint classification
binary image
core
ridge following
