摘要
本文就视网膜血管形态生物信息的自识别方法所展开的初步探索，介绍相关研究方法的状况，提出以最大根部网孔表征和提取其形态特征，由网孔拓扑关系建立弹性拓朴匹配模式，实现与网孔方向特征定位无关的形态匹配以及利用多级判据措施提高识别率等方法问题。
This paper describes a preliminary research of self-recognition approach for the biology information of vascular pattern of retina, presents the situation of relating research methods, advances the theoretic approaches that the largest root network can be used to represent and extract.the features of vascular Pattern of retina,and that the elastic topology matching model can be set up through the topology relation of network so that the pattern matching is without relating to the initial orientation of oriented features of network,as well as that the recongition probability can be enhanced by means of multi-cascade criterion.
出处
《华南理工大学学报（自然科学版）》
EI
CAS
CSCD
1995年第8期121-125,共5页
Journal of South China University of Technology(Natural Science Edition)
基金
国家自然科学基金
关键词
视网膜
血管
模式识别
自识别
形态
生物信息
s:retinal vessels
pattern recognition
topological features/root network
elastic topology matching model
self-recognition
