| 金梅,薛静芳,张立国,刘强.基于PERCLOS判据的驾驶员疲劳监测系统[J].高技术通讯(中文),2022,32(12):1245~1250 |
| 基于PERCLOS判据的驾驶员疲劳监测系统 |
| Driver fatigue monitoring system based on PERCLOS criterion |
| |
| DOI：10. 3772/ j. issn. 1002-0470. 2022. 12. 004 |
| 中文关键词: 疲劳监测； 流水线算法； 人脸检测； 人眼跟踪； 单位时间内人眼闭合时间所占的比例（PERCLOS）判据 |
| 英文关键词: fatigue monitoring, pipeline algorithm, face detection, eyetracking, percentage of eyelid closure over the pupil over time （PERCLOS） criterion |
| 基金项目: |
| 作者 | 单位 | | 金梅 | （燕山大学电气工程学院秦皇岛 066000） | | 薛静芳 | （燕山大学电气工程学院秦皇岛 066000） | | 张立国 | （燕山大学电气工程学院秦皇岛 066000） | | 刘强 | （燕山大学电气工程学院秦皇岛 066000） |
|
| 摘要点击次数: 738 |
| 全文下载次数: 541 |
| 中文摘要: |
| 针对疲劳驾驶监测常用算法运算效率低的问题，提出了一种基于单位时间内人眼闭合时间所占比例（PERCLOS）判据的驾驶员疲劳监测方案。针对传统顺序执行算法运算效率低的问题，提出了基于流水线算法的人脸检测及人眼跟踪算法。针对传统顺序式算法对硬件要求高的问题，本系统运行在低端现场可编程门阵列（FPGA）上，设备成本大幅降低。并利用PERCLOS判据作为判断驾驶员是否疲劳的依据。实验表明，本文所提出的基于流水线算法的图像处理算法使系统拥有了较高的实时性和较低的资源占用，实现了60fps的处理速度，系统的准确度达到了95%以上。 |
| 英文摘要: |
| Aiming at the low efficiency of commonly used algorithms for fatigue monitoring, this paper proposes a driver fatigue monitoring program based on the percentage of eyelid closure over the pupil over time （PERCLOS） criterion. Aiming at the low efficiency of traditional sequential execution algorithms, a face detection and eye tracking algorithm based on pipeline algorithm is proposed. Aiming at the problem of high hardware requirements for traditional sequential algorithms, this system runs on low end field programmable gate array （FPGA）, and the equipment cost is greatly reduced. The PERCLOS criterion is used as the basis for judging whether the driver is fatigued. Experimental results show that the image processing algorithm based on the pipeline algorithm proposed in this paper enables the system to have higher real time performance and lower resource consumption. A processing speed of 60fps has been achieved, and the test results show that the accuracy of the system has reached more than 95%. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
