模式识别和电视跟踪在前庭功能检查中的应用

Application of Mode Recognition and TV Tracking for Checking the Vestibula Function

应用模式识别和电视跟踪原理测量眼球在视靶信号刺激下的运动,为分析飞行员、车辆驾驶员的前庭功能提供一种先进的检查方法。采用均衡化、形态滤波和形心计算等方法实现了瞳孔区域的自动定位和实时跟踪;通过Harris角点检测和Lucas-Kanade光流跟踪实时测量出虹膜的旋转角度,为前庭功能分析奠定了良好的基础。通过实验验证,所介绍的眼球跟踪算法已达到了预期的效果,并已成功应用在前庭功能检查分析系统中。

An advanced examination-method is brought forward aiming at analyzing the vestibular function of pilots and car-drivers. It surveys eyeball movement by using of mode recognition and TV tracking theory when the irritant signals act on the eyes. R adopts some advanced methods such as equalization, morphological filtering, and the centroid tracking which carry out auto-localization and real-time tracking of pupil area. It can measure circumgyrated angle of the iris by using of Harris comer detecting and Lucas-Kanade Optical Flow tracking, and gives a better foundation for vestibular function analysis. Experiments validate eyeball tracking algorithm achieve better results, and have put it to the practice of vestibular function checking and analysis system.