摘要
基于机器视觉的疲劳检测难点是人眼的准确、自动化的定位。针对目前基于红眼效应的人眼定位算法需要手工介入的问题,提出了一种改进算法,采用眼部区域的几何特征和对比度相结合的算法实现了自动化人眼定位,形成了一种基于红眼效应的实时、自动化定位人眼的快速算法。然后构建了一套使用850 nm红外光源的人眼定位系统,实现了快速准确的定位人眼。
The difficulty of fatigue-detecting based on machine vision is to locate the eyes automatically. At present, the eyes location algorithm based on pinkeye allocation needs artificial intervention. In this article an improved algorithm was presented, which applies geometry characters and contrast to the automatic eyes location. Thus, an advanced real-time algorithm was formed to locate the eyes automatically. At last, an eyes locating system was built, which used 850nm infrared light to locate the eyes quickly and precisely.
出处
《交通与计算机》
2007年第6期34-36,40,共4页
Computer and Communications
关键词
红外光源
人眼定位
几何特征
对比度
infrared light
eyes location
geometry characters
contrast
