|
| 驾驶防瞌睡装置中人眼快速定位方法研究 |
| Fast location of eye features in drowsy driving warning system |
| |
| DOI: |
| 中文关键词: 积分投影,边缘检测,特征提取 |
| 英文关键词:integral projection,edge detection,feature extraction |
| 基金项目: |
| 李峰 曾超 徐向东 |
浙江大学现代光学仪器国家重点实验室 浙江杭州310027
(李峰,曾超) ,浙江大学现代光学仪器国家重点实验室 浙江杭州310027(徐向东)
|
| 摘要点击次数: 2019 |
| 全文下载次数: 11 |
| 中文摘要: |
| 驾驶防瞌睡装置用于实时监测司机眼皮活动情况 ,介绍了用于该装置基于灰度积分投影的人眼快速定位方法。首先利用图像的垂直灰度投影曲线确定人脸左右边界 ,然后利用水平灰度投影曲线确定人眼位置 ,最后利用阈值分割及边缘检测进行人眼开闭情况检测。实验表明 ,该方法对于驾驶防瞌睡装置中的人眼位置判定具有很强的实用价值 |
| 英文摘要: |
| Drowsy driving warning system is used in real time watching movements of driver's eyes.In this paper a method of fast locating eye features in facial images obtained from the system is introduced.First,the left and right boundaries of face are localized with vertical gray level projection.Second,the eye region is localized based on horizontal gray level projection.Finally,threshold segmentation and edge detection method are adopted to determine the state of eyes.Experiment results show that the method has a high accuracy and robust performance for our drowsy driving warning system. |
| HTML 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|