用于精神分裂症检测的眼动仪研制

An eye tracker used for the detection of schizophrenia

目的为了辅助医生诊断精神分裂症患者,设计一种高精度桌面式眼动仪。方法系统采用双屏输出,摄像机集成了红外光源及红外滤光镜片。重点提出一种新的瞳孔中心识别算法,首先对预处理后的眼图进行瞳孔轮廓提取,然后采用最小二乘椭圆拟合法识别瞳孔中心。在此基础上,对瞳孔周围区域进行边缘检测,得到的最大轮廓中心即为普尔钦光斑中心。最后基于非线性二阶多项式拟合静态标定算法建立了眼图坐标系和场景坐标系的映射方程,并完成了注视点标定。结果实现了眼动轨迹的实时显示和凝视点数目的统计,测试有效率达到80%以上。结论系统能够很好地跟踪和显示人眼运动,具有较好的实用性。

Objective In order to help doctors to diagnose schizophrenia patients,a desktop eye tracker with high-precision was designed in this paper. Methods The system used dual output,and the camera is integrated with infrared light source and infrared filter lens. A new recognition algorithm for pupil center was proposed. The pupil contour was extracted after preprocessing the imagine,then the pupil center is identified based on the least squares ellipse fitting algorithm. On this basis,the center of the largest contour that was achieved by edge detection around the pupil center was Purkinje spot. Finally,the static calibration algorithm of the nonlinear second-order polynomial fitting established the mapping equation between eye coordinate system and scene coordinate system,and the fixation point calibration was completed. Results The system realized the real-time display of the eye movement trajectory and the statistics of the fixation points,and the effective rate of the test was above 80%. Conclusions The eye tracker developed in this paper can track and display the human eye movement very well,and has good practicability.