针对目前已有的视点跟踪算法普遍存在的算法复杂度较高,难以实现实时性问题,提出了一种低复杂度的快速视点跟踪方法,先捕获红外光源下的人脸图像,使用高低两次阈值处理后,通过8个方向的游程长来检验红外反射点,从而定位瞳孔,之后利用与瞳孔-反射点距离向量的映射关系来确定用户在屏幕上的注视点坐标.结果表明,该方法在在保证精度的同时,满足了高分辨率下实时性处理的要求.
As existing gaze tracking methods are too complex to be run in real-time, a new fast method under low complexity is presented. Process face images under infrared lamp-houses with two valves, and verify pupil area by the run-length in 8 directions, then calculate the gaze point by mapping to the vectors of pupil and reflection point. The experimental results show that the real-time processing of gaze tracking, with high resolution and high precision simultaneously, is achieved in this new system.