一种红外图像条件下的驾驶人眼睛定位算法

Algorithm for locating driver's eyes under conditions of infrared images

利用机器视觉对驾驶人面部器官状态进行监测和分析处理是安全辅助驾驶领域内的研究热点之一。首先介绍了驾驶人驾驶行为监测及预警系统的原理和结构,然后讨论了系统中驾驶人眼睛定位的技术要点。通过人类眼睛虹膜对不同波长的红外光吸收能力的不同,利用特制的红外光源获取到驾驶人面部的两种红外反射图像,利用图像处理算法进行处理从而获得驾驶人的眼睛位置,从而为下一步的眼睛状态及信息提取打下了良好的基础。实践证明,算法的实时性高,效果非常理想。

Monitoring and analysing the state of driver’s facial organs by using the machine vision is one of the hotspots in the assistant safe driving system.This paper introduces the theory and structure of monitoring and warning system for driver’s driving behavior firstly,and then discusses the technical points in locating driver’s eyes in this system.In respect that the iris of a human being reflects and absorbs infrared rays at different wavelength differently,with special light sources of infrared ray,the author obtain different infrared reflected images,hereby catch the location of driver’s eyes by using the image-processing algorithm,which lay a good foundation for further information-collecting and state-analysing of driver’s eyes.Practice proves that the algorithm used in this paper acts in realtime and obtains excellent effect.