基于非接触式眼压计的双目定位方法研究  

Research on Binocular Positioning System Based on Non-contact Tonometer

作　　者：王苗 石利霞[1] 杨成禹[1] 余晨洲 WANG Miao;SHI Li-xia;YANG Cheng-yu;YU Chen-zhou(School of Optoelectronic Engineering,Changchun University of Science and Technology,Changchun 130022)

机构地区：[1]长春理工大学光电工程学院,长春130022

出　　处：《长春理工大学学报（自然科学版）》2020年第2期63-67,共5页Journal of Changchun University of Science and Technology(Natural Science Edition)

基　　金：吉林省重点科技研发项目(20180201025GX)。

摘　　要：基于非接触式眼压计对人眼定位系统进行研究,采用汇聚式双目视觉模型对人眼进行定位。阐释了双目视觉原理和相机标定原理与流程,通过标定评价方法对标定结果进行评价,同时搭建双目视觉系统实验平台,对真实人眼进行试验。对眼图处理过程进行了分析,采用HOUGH圆检测与SURF-RANSAC算法进行特征点匹配,最后得到空间点与三维定位点坐标间误差,实验结果验证了方案可行性。In this paper,the human eye positioning system based on the non-contact tonometer was studied;and the convergent binocular vision model was used to locate the human eye. The principle of binocular vision and the principle and process of camera calibration were explained. The calibration results were evaluated by the calibration evaluation method. At the same time,the binocular vision system experimental platform was built to test the real human eye. The eye diagram processing process was analyzed. The HOUGH circle detection and SURF-RANSAC algorithm were used to perform feature point matching. Finally,the error between the spatial point and the 3 D positioning point coordinates was obtained. Experimental results verify the feasibility of the program.

关 键 词：非接触式眼压计 人眼定位 双目视觉 相机标定 MATLAB 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]