红外头盔式眼动仪的瞳孔中心定位算法

doi:10.3969/j.issn.1001-2400.2011.03.002

J4 ›› 2011, Vol. 38 ›› Issue (3): 7-12+82.doi: 10.3969/j.issn.1001-2400.2011.03.002

红外头盔式眼动仪的瞳孔中心定位算法

王军宁；刘涛；何迪；武媛媛

(西安电子科技大学通信工程学院，陕西西安 710071)

收稿日期:2010-05-18 出版日期:2011-06-20 发布日期:2011-07-14
通讯作者: 王军宁
作者简介:王军宁(1958-)，男，教授，E-mail: xdtidsp@mail.xidian.edu.cn．
基金资助:
高等学校学科创新引智计划资助项目(B08038)

Pupil center localization algorithm used for the IR head-mounted eye tracker

WANG Junning；LIU Tao；HE Di；WU Yuanyuan

(School of Telecommunication Engineering, Xidian Univ., Xi'an 710071， China)

Received:2010-05-18 Online:2011-06-20 Published:2011-07-14
Contact: WANG Junning

摘要/Abstract

摘要：

为了提高瞳孔中心定位精度，降低运算复杂度，首先采用星射线方法获取瞳孔边界点，与对整幅图搜索相比，计算量大大减小．设定感兴趣区域，对眼皮、眼睫毛及光斑干扰产生的虚假特征点进行剔除，然后利用特征点通过随机化的椭圆拟合定位瞳孔中心．随机化椭圆拟合允许杂质点的存在，确保了拟合定位算法的准确性，对虚假点的剔除更进一步加快了拟合速度．在连续帧处理时通过瞳孔像素数的变化快速对眨眼进行检测，区分眨眼与伪眨眼．实验结果表明，该算法简单且鲁棒性好．

关键词: 瞳孔中心定位, 星射线, 椭圆拟合, 眨眼检测

Abstract:

In order to improve pupil center localization accuracy and reduce computational complexity, firstly, we adopt star rays to get pupil counter points. Its amount of calculation is cut down drastically compared with searching in the whole figure. Secondly, we cope with some false points produced by eyelashes, eyelids and glints in the interesting area. Then, we apply a randomized method of ellipse fitting to determine the pupil center with pupil counter points. The inclusion is allowed in randomized ellipse fitting so as to ensure the accuracy of the fitting localization algorithm. And eliminating false points further increases the fitting rate. The variation of the number of pupil pixels is detected to distinguish between true-blink and pseudo-blink in dealing with continuous frames. Experiments show that the new algorithm is simple and robust.

Key words: pupil center localization, star rays, ellipse fitting, blink detection

王军宁；刘涛；何迪；武媛媛. 红外头盔式眼动仪的瞳孔中心定位算法[J]. J4, 2011, 38(3): 7-12+82.

WANG Junning；LIU Tao；HE Di；WU Yuanyuan. Pupil center localization algorithm used for the IR head-mounted eye tracker[J]. J4, 2011, 38(3): 7-12+82.

参考文献

［1］ Morimoto C H, Mimica M R M. Eye Gaze Tracking Techniques for Interactive Applications ［J］. CVIU, 2005, 98(1): 4-24.
［2］刘伟, 袁修干. 人的视觉-眼动系统的研究［J］. 人类工效学, 2000, 6(4): 41-44.
Liu Wei, Yuan Xiugan. Research on the Eyes System of Human Vision ［J］. Chinese Ergonomics, 2000, 6(4): 41-44.
［3］武媛媛. 眼动跟踪技术研究［D］. 西安: 西安电子科技大学, 2009.
［4］ Li D H, Win F D, Parkhurst D J. Starburst: a Hybrid Algorithm For Video-based Eye Tracking Combining Feature-based and Model-based Approaches ［EB/OL］. ［2010-01-15］. http://hcvl. hci. iastate. edu/openEyes.
［5］张仁彦, 赵国良. 一种基于区域标记的虹膜内边缘定位新算法［J］. 海军工程大学学报, 2005, 17(4): 18-21.
Zhang Renyan, Zhao Guoliang. A New Algorithm For Location of Iris Inner Boundary Based on Region Marking ［J］. Journal of Naval University of Engineering, 2005, 17(4): 18-21.
［6］闫蓓, 王斌, 李媛. 基于最小二乘法的椭圆拟合改进算法［J］. 北京航天航空大学学报, 2008, 34(3): 295-298.
Yan Bei, Wang Bin, Li Yuan. Optimal Ellipse Fitting Method Based on Least-square Principle［J］. Journal of Beijing University of Aeronautics and Astronautics, 2008, 34(3): 295-298.
［7］ Ko Y J, Lee E C, Park K R. A Robust Gaze Detection Method by Compensating for Facial Movements Based on Corneal Specularities ［J］. Pattern Recognition Letters, 2008, 29(10): 1474-1485.
［8］刘瑞安, 靳世久, 李文清, 等. 瞳孔亚像素边缘检测与中心定位［J］. 计算机工程与应用, 2007, 43(5): 200-203.
Liu Rui'an, Jin Shijiu, Li Weiqing, et al. Subpixel Edge Detection and Center Localization of The Pupil ［J］. Computer Engineer and Applications, 2007, 43(5): 200-203.
［9］苑玮琪, 马军防, 狄文彬, 等. 基于主动轮廓线的虹膜内边界的定位方法［J］. 红外与激光工程, 2003, 32(6): 605-609.
Yuan Weiqi, Ma Junfang, Di Wenbin, et al. New Method of Interior Boundary of Iris Location Based on Active Contour ［J］. Infrared and Laser Engineering, 2003, 32(6): 605-609.
［10］冯成志, 沈模卫. 视线跟踪技术及其在人机交互中的应用［J］. 浙江大学学报, 2002, 29(2): 225-232.
Feng Chengzhi, Shen Mowei. Application of Tracking in Human Computer Interaction ［J］. Journal of Zhejiang University, 2002, 29(2): 225-232.

红外头盔式眼动仪的瞳孔中心定位算法

Pupil center localization algorithm used for the IR head-mounted eye tracker

PDF (PC)

赞

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

Metrics

本文评价

推荐阅读 10