飞行员疲劳监测中的人眼定位及状态判别研究

摘要

飞行员驾驶疲劳是影响飞行安全的重要因素之一。常用基于人眼识别的方法对飞行员进行疲劳监测，其关键是人眼定位及状态判别。利用改进的平均合成滤波器算法定位人眼，根据模糊综合评价思想，判定人眼睁开闭合状态，根据PERCLOS疲劳检测方法对飞行员进行疲劳监测。实验表明，改进的人眼定位方法具有更高的精确度和鲁棒性，提出的人眼状态判定方法能够满足疲劳监测需要。为最终构建一种机载的、实时的、非接触式的飞行员疲劳监测系统提供了理论基础。%Pilot fatigue is one of the important factors that affect flight safety. We usually monitor pilot fa-tigue by eyes. The key is how to locate eyes and recognize the state of eyes. In this paper,we firstly use the improved average synthetic exact filter algorithm to locate the eyes. Then,we judge whether the eyes are open or closed according to fuzzy comprehensive evaluation method. Finally,we monitor the pilot′s fa-tigue state by PERCLOS fatigue detection method. The experiments show that the improved eye location algorithm is with higher accuracy and robustness,and the judgement method of eye state can satisfiy the requirement of fatigue monitoring. The theoretical basis is provided for the final construction of an air-borne,real-time,non-contact pilot fatigue monitoring system.