基于眼角精确定位的眼睛状态识别方法
首发时间:2012-03-20
摘要:眼睛的状态识别是疲劳研究的重要内容之一。本文提出了基于特征分析的一种新颖的眼睛定位状态识别方法。首先采用adaboost 中的haar人脸特征检测方法对图像进行人脸检测,其次运用一种新的眼角定位模型定位眼睛,然后根据眼角位置定位出眼睛状态匹配区域,运用模板匹配的方法进行眼睛状态识别,实验表明该方法具可行性和鲁棒性。
关键词: 状态识别 眼角定位 模板匹配 轮廓提取 椭圆拟合 角点检测
For information in English, please click here
Eye state recognition based on the Accurate Eye corners location
Abstract:Eye corners location and eye state recognition is important content to monitor the driver's fatigue state. A novel eye corner localization and eye state recognition method is proposed in the paper. Firstly, a haar features face detection based on adaboost is used to detect the face in an image. Secondly, a step of accurate eye corner location is proposed. Then the matching area of eye state is located by the eye corners location. The match method which using the template carries on the eye condition recognition. The proposed method is proved to be an effective and robust method according to the result of comparative experiments.
Keywords: state recognition eye corner localization template-matching contour extraction ellipse fitting corner detection
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
基于眼角精确定位的眼睛状态识别方法
评论
全部评论0/1000