引用本文: |
穆珺,王勇,晏峻峰,彭清华.面向中医目诊的虹膜斑块定位算法[J].湖南中医药大学学报,2018,38(10):1149-1153[点击复制] |
|
|
|
本文已被:浏览 2690次 下载 1378次 |
面向中医目诊的虹膜斑块定位算法 |
穆珺,王勇,晏峻峰,彭清华 |
(湖南中医药大学, 湖南 长沙 410208) |
摘要: |
目的 研究和实现面向中医目诊的虹膜斑块定位,以利于深入研究虹膜及其图像特征与疾病、体质等因素间的对应关系。方法 以虹膜仪采集图像为对象,研究并提出了一种基于中医目诊的虹膜斑块定位算法。该算法首先对虹膜图像预处理;再进行虹膜图像的区域对比特征计算;最后基于区域对比特征,通过阈值化和联通区域几何特征,进行虹膜斑块定位。结果 本文算法在自行采集的虹膜仪图像数据集上进行测试,定位的精度(precision)为74.3%,召回率(recall)为83.9%。本文所提算法的运行时间,约为0.312 min每张虹膜图像。结论 本文方法能有效定位虹膜斑块,定位准确度和运行速度优于现有方法。 |
关键词: 中医目诊 虹膜斑块定位 区域对比特征 积分图 |
DOI:10.3969/j.issn.1674-070X.2018.10.014 |
投稿时间:2018-01-04 |
基金项目:十三五国家重点研发计划课题资助(2017YFC1703306);中医药防治五官科疾病湖南省重点实验室建设项目(2017TP1018);长沙市科技计划项目(kc1704005);中央财政支持地方高校建设项目;国家中医药管理局中医药信息学重点学科建设项目;国家中医药管理局中医眼科学重点学科建设项目;2013年湖南省高校创新平台开放基金(13K076);国家重点学科中医诊断学开放基金(2013ZYZD08);2016年校级大学生研究性学习和创新性实验计划课题;湖南中医药大学校级青年教师科研基金课题。 |
|
An Iris Plaque Localization Algorithm for Traditional Chinese Medicine Eye Diagnosis |
MU Jun,WANG Yong,YAN Junfeng,PENG Qinghua |
(Hunan University of Chinese Medicine, Changsha, Hunan 410208, China) |
Abstract: |
Objective To present an iris plaque localization algorithm for traditional Chinese medicine (TCM) eye diagnosis, and to further study the relationship of iris image characteristics with disease and constitution. Methods An iris plaque localization algorithm was proposed for TCM eye diagnosis based on iris images. The algorithm computed the regional contrast characteristics of iris images after preprocessing. Finally, according to the regional contrast characteristics, iris plaque was located by thresholding and linking regional geometric features. Results The proposed algorithm was tested in an independent iris image dataset, yielding a localization precision of 74.3% and a recall rate of 83.9%. The proposed algorithm took around 0.312 minute to process one iris image. Conclusion The proposed algorithm can effectively locate the iris plaque, which has superior accuracy and speed to existing algorithms. |
Key words: traditional Chinese medicine eye diagnosis iris plaque localization regional contrast characteristics integral map |
|
二维码(扫一下试试看!) |
|