基于ZYNQ MPSOC的驾驶员面部疲劳检测设计  

On Driver Face Fatigue Detection Based on ZYNQ MPSOC

作　　者：雷小英 肖顺文 王涌 吴静瑜 王睿 LEI Xiaoying;XIAO Shunwen;WANG yong;WU Jingyu;WANG rui(School of Physics and Astronomy,China West Normal University,Nanchong Sichuan 637009,China;School of Electronics and Information Engineering,China West Normal University,Nanchong Sichuan 637009,China)

机构地区：[1]西华师范大学物理与天文学院,四川南充637009 [2]西华师范大学电子信息工程学院,四川南充637009

出　　处：《乐山师范学院学报》2023年第4期50-56,共7页Journal of Leshan Normal University

基　　金：校科研创新团队基金项目(CXTD2017-8);校英才科研基金项目(17YC055)。

摘　　要：针对驾驶员疲劳驾驶车辆易致重大交通事故,设计了基于驾驶员面部特征疲劳检测系统。在ZYNQ平台下,利用OV5640摄像头进行图像采集,在PL端进行图像处理,在PS端的ARM中显示图像。OpenCV调用训练的人脸分类器进行人脸检测,通过分类器得到人脸图像,再由三庭五眼的特征分割出眼睛和嘴巴区域,后对分割后的图像进行图像阈值化处理和膨胀得到眼睛和嘴巴的大致轮廓。通过对分割后的眼睛和嘴巴的外接矩形的宽长比进行计算,结合PERCLOS算法、眨眼及打哈欠频率快速侦察驾驶员是否出现疲劳。在实验室模拟驾驶环境进行测试,该系统能实时准确检测到驾驶员面部疲劳状态。A fatigue detection system based on driver's facial features is designed to prevent the serious traffic accidents caused by driver's fatigue driving.Under the ZYNQ platform,the OV5640 camera is used for image acquisition,image processing is carried out at the PL end,and the image is displayed in the ARM at the PS end.OpenCV calls the trained face classifier for face detection.The face image is obtained through the classifier,and then the eye and mouth regions are segmented from the characteristics of‘‘three chambers and five eyes''.Then the segmented image is processed by threshold and expanded to obtain the approximate contour of eyes and mouth.By calculating the width length ratio of the circumscribed rectangle of the segmented eyes and mouth,combining with PERCLOS algorithm and blinking and yawning frequency,we can quickly detect whether the driver is tired.Tested in the simulated driving environment in the laboratory,the system can detect the driver's facial fatigue state in real time and accurately.

关 键 词：疲劳检测 ZYNQ OPENCV PERCLOS 

分 类 号：TN02[电子电信—物理电子学]