一种基于图像处理的人眼全自动角膜曲率计研究  

Auto-ophthalmometer of Eye Based on Image Processing

作　　者：赵俊奇[1] 郭智勇[1] 陈安世[2] 刘海峰[1] 

机构地区：[1]中北大学光电仪器厂,太原030051 [2]清华大学环境工程系,北京100084

出　　处：《中国生物医学工程学报》2011年第1期100-104,共5页Chinese Journal of Biomedical Engineering

基　　金：科技部科技成果重点推广项目(2003EC000084);太原市科技创业种子基金(082114)

摘　　要：提出一种以光学系统成像原理的自动角膜曲率计设计方案,并研制完成实验样机,对样机的检测原理、系统工作原理进行说明。该研究方法是将红外园环照到眼睛角膜,经角膜反射成像在CCD上,首先局部进行边缘检测,求取全局阈值算法进行二值化图像,然后检测出二值化边缘轮廓,最后再用最小二乘拟合的方法,计算测量角膜的曲率参数。用标准角膜模拟眼和人眼进行测试,测量精度达到±0.02 mm,测量时间小于1 s。实验结果表明,该方法可行,可以用于医院等角膜曲率参数测量的场合。An auto-ophthalmometer based on optical imaging was developed and tested.The principles for detection and operation of the system were presented.Infrared image circle reflected from retina was received by CCD.At first,local edge detection was performed to get the threshold for the binaryzation of the whole image.Then,the edge was extracted and the diopter parameter of the retina was calculated through least squares fitting.The experiments on simulation eyes and human eyes showed that the precision of the measurement could reach ± 0.02 mm with time consuming less than 1 s.The system is feasible and can improve the measurement accuracy of the diopter parameter.

关 键 词：角膜曲率计 DSP系统 边缘检测 

分 类 号：R318[医药卫生—生物医学工程]