摘要
据交通部门对道路交通事故死亡人数的统计，我国连续数年一直居世界第一位，汽车司机的疲劳驾驶是引发交通事故的一个重要原因。因此对驾驶疲劳检测以及如何有效的监测和防止驾驶员疲劳驾驶的研究，对于降低交通事故，有着十分重要的现实意义。
本文首先介绍了国内外有关驾驶疲劳的研究现状，重点研究了眼睛的疲劳机理，借鉴美国联邦公路管理局FHWA(Federal HighwayAdministration)有关研究驾驶疲劳方面的文献，讨论了PERCLOS(percentage 0f eyelid closure over the pupil over time单位时间内眼睛闭合时间所占的百分率)和其他眼睛活动测量方法的有效性，从理论上证明PERCLOS是目前测量机动车驾驶疲劳最佳方法，因此本文中选用PERCLOS作为驾驶员疲劳程度判定的依据。
其次，本文详细分析了当前眼睛疲劳检测方法中的技术关键和难点，在数字图像处理基础上完成人脸实时检测及人眼定位跟踪算法，引入灰色理论作为图像处理的手段，对灰色系统理论的基本思想和一些重要的定义、定理进行了较深入地介绍和推导，介绍了基于灰色预测模型方法和基于灰色预测模型的非线性图像滤波器设计方法。利用AGO数据生成和GM(1，1)模型，设计出一套滤波算法和基于灰色关联的边缘检测算法，并进行了理论证明和MATLAB实现。
最后针对疲劳检测算法中大数据量、高速传输、复杂运算的实际需要，构建了由CCD、TMS320C6416、SAA7115、SAA7105等芯片组成的视频采集系统。设计了一种车载的、非接触式的、实时的、基于DSP实时驾驶疲劳检测系统。
According to the Department of Transportation for road traffic accident mortality statistics, China has for several consecutive years ranked No. 1. Driver fatigue driving is a traffic accident caused by an important reason. Therefore driving fatigue detection and effective ways to monitor and prevent driver fatigue driving, reducing traffic accidents, have great practical significance.
This paper introduces the driving fatigue of the status and focus on the eye fatigue mechanism, fromtheFederal Highway Authority.(Federal HighwayAdministra tion) The study of fatigue driving literature, discussed PERCLOS (percentage of eyelid closure over the pupil over time) and other eye - dynamic measurement of the effectiveness, proved theoretically PERCLOS measurement is the best vehicle driving fatigue, This paper therefore choose PERCLOS driver fatigue as a basis for judgment.
Secondly, the paper detailed analysis of the current eye fatigue detection method of key technical and difficult, Digital image processing completed on the basis of real-time face detection and tracking eye positioning algorithm, Gray introduced as a means of image processing, gray system theory and the basic idea of some important definitions, theorem for a more in-depth presentations and deduced, Based on the forecast model gray and gray forecasting model based on the non-linear image filter design methods. Use AGO data generation and GM (1,1) model, design a filter algorithm and associated gray edge detection algorithm, and the proof of theory and simulation.
Finally fatigue detection algorithm large data volume, high-speed transmission, complex operation to the actual needs and building a CCD, TMS320C6416, SAA7115, SAA7105 chip components such as the Video Acquisition System. Design of a vehicle, non-contact, Real-time driving fatigue detection system-DSP-based real-time driving fatigue detection system.
引文
[1]王荣本，郭克友．驾驶员驾驶行为监测中的面部定位方法的研究．公路交通科技，2003(2)：33-35
[2]阎淑芳．机动车驾驶员操作技能分析．人类工效学，1997(2)：60-61
[3]石坚，卓斌．汽车驾驶员主动安全性因素的辨识与分析．上海交通大学学报，2000(4)：441-444
[4]Kawato S, Tetsutani N. Real-time Detection of Between-the-Eyes with a Circle Frequency Filter. The 5 Asian Conference Computer Vision. Melbourne, Australia, 2002(1): 23-25
[5]Ellen M.A, Grace.R, Steinfeld. A User-Centered Drowsy-Driver Detection and Warning System, Proceedings of ACM Designing User Experiences, 2003(5): 109-113
[6]Horng WB, Chert CY, Chang Y, Fan CH. Driver fatigue detection based on eye tracking and dynamic, template matching. IEEE International Conference on Networking, Sensing and Control. 2004(10): 7-12
[7]Gu HS, Ji Q. An automated face reader for fatigue detection. Sixth IEEE International Conference on Automatic Face and Gesture Recognition, 2004:111-116
[8]左坤隆，刘文耀．基于活动外观模型的人脸表情分析与识别．光电子·激光，2004(7)：853-857
[9]郑培，周一鸣．机动车驾驶员驾驶疲劳测评方法的研究状况及发展趋势．中国农业大学学报，2001(6)：53-55
[10]郭克友，储江伟，王荣本．驾驶员眼部状态识别方法的研究．计算机学报，2003(10)：1186-1188
[11]章毓晋．图像处理和分析．北京：清华大学出版社，2002
[12]张杰，杨晓飞，赵瑞莲．基于Hough变换圆检测的人眼精确定位方法．计算机工程与用，2005(27)：43-46
[13]于威威，滕晓龙，刘重庆．复杂背景下人眼定位及人脸检测．计算机仿真，2004(12)：185-188
[14]尹方平，赖剑煌，阮邦志．非均匀光照下人脸眼睛的定位方法．中山大学学报(自然科学版)，2003(5)：12-15
[15]钟威，刘智明，周激流．人脸检测中眼睛精确定位的研究．计算机工程与应用．，2004(36)：73-77
[16]史慧荣，张学帅，梁彦等．一种基于模糊分类的模板匹配眼睛定位方法，2005(2)：55-59
[17]艾海舟，肖习攀，徐光枯．人脸检测与检索[J]．计算机学报，2003 26(7)：874-881
[18]梁路宏，艾海舟，徐光祷等．人脸检测研究综述．计算机学报，2002 25(6)：449-458
[19]Albert P, Jordi Vitrib. Topological Principal Component Analysis for Face Encoding and Recognitionr[J]. Pattern Recognition letters, 2001 22(627): 7692-7761
[20]Govindaraju V, Srihari S N, Sher D B. A computational model for face location[J]. Computer Vision, 1990 23(5): 718-721
[21]林峰，张政凯，愈能海．基于图象分割的人眼边缘提取方法研究．计算机应用研究，2000(8)：100-101
[22]Kim H, Kang W, Shin J, Park S. Face Detection Using Template Matching and Ellipse Fitting[J]. IEEE Tranc. 2000(11): 2008-2011
[23]Wang J G, Tan T N. A new face detection method based on shape information[J]. Pattern RecoQnition, 2000 21(67): 463-471
[24]I Craw, H Ellis, J Lishman. Automatic extraction of face features[J]. Pattern Recognition, 1987 5(2): 183-187
[25]Yoo T.W, Oh I.S. A fast algorithm for tracking human faces based on chromatic histograms[J]. PatternRecognition, 1999 20(10): 967-978
[26]Wei G, Sethi I K. Face detection for image annotation[J]. Pattern Recognition, 1999 20(11): 1313-1321
[27]Abdel-Mottaleb M, Elgammal.A. Face detection in complete environments from color images[J]. ImageProcessing, 1999 3(113): 622-626
[28]Karlekar J, Desai U B. Finding faces in color images using wavelet transform[J]. Image Analysis andProcessing, 1999 13(36): 1085-1088
[29]章毓晋．图像工程(上、下)．北京：清华大学出版社．2002
[30]郑南宁．计算机视觉与模式识别．北京：国防工业出版社．1998
[31]Douglas Chai. Face Segmentation Using Skin-color Map in Video Phone Applications. IEEE Trans.CSVT. 1999 9(4): 551-564
[32]韩相军，关永，王雪立．基于DSP的疲劳驾驶实时监测系统研究．微机发展，2006 16(2)：47-49
[33]Lam K, Yan H. Locating and extracting the eye in human face images[J]. Pattern Recognition, 1996 29(5): 771-779
[34]Antlonio Haro, Myron Fickner, Irfan Essa. Detecting and tracking eyes by using their physiological properties,dynamics and appearance[C]. France. 2000(9): 234-237
[35]L Huang, CW Chen. Human Facial Feature Extraction for Face Interpretation and Recognition[J]. Pattern Recognition, 1992 25(12): 1435-1444
[36]X Xie, R Sudhakar, H Zhuang.On Improving Eye Feature Extraction Using Deformable Templates[J]. Pattern Recognition, 1994 27(6): 791-799
[37]T Morris, P Blenkhorn, Farhan Zaidi. Blink Detection for Real-Time Eye Tracking[J]. Net and Computer annlications, 2002 37(2): 129-143
[38]李英，赖剑煌，阮邦志．眼睛状态检测的组合方法[J]．中国图象图形学报，2003 8(10)：1140-1145
[39]Chow Q, Li X. Towards A system for Automatic Facial Feature Detection[J]. Pattern Recognition 1993 26(12): 1739-1755
[40]Tian Y L, Kanade T, Cohn J E Recognizing upper face actions for facial expression analysis[A]. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR'00)[C], Hilton Head Island, South Carolina, USA. 2000(1): 294-301
[41]韩相军，关永，王万森，王雪立．基于DSP的嵌入式驾驶疲劳监测系统研究[J]．公路交通科技，2006(12)：56-59
[42]胡相军．基于DSP的驾驶疲劳实时检测系统研究．计算机工程与科学，2000(3)：78-83
[43]吴国平．灰色系统理论与矿床灰色预测．武汉：中国地质大学出版社．1993
[44]周卫．基于MATLAB的灰色系统沉降预测．测绘通报，2002(6)：34-36
[45]毛弋，杨期余，刘义仁，刘文春．灰色预测模型的有限自动机的实现问题．湖南大学学报(自然科学版)，2001 28(1)：64-68
[46]彭春，吴国平，尹兰．一种快速灰色预测模型．现代电子技术，2002 33(2)：45-47
[47]邓聚龙．灰色控制系统(第二版)．武汉：华中理工大学出版社．1993
[48]顾晓辉，王晓鸣，赵有守．基于灰色系统预测直升机航迹的研究．航空学报，2001 22(6)：559-562
[49]马苗等．基于灰色系统理论的图象边缘检测新算法．中国图象图形学报，2003(10)：1136-1139
[50]郑子华．基于灰色绝对关联度的边缘检测．福建师范大学学报(自然科学版)，2004 1(12)：20-23．
[51]郑培，周一鸣．基于PERCLOS的机动车驾驶员驾驶疲劳的识别算法．中国农业大学学报，2000(6)：33-36
[52]王念旭．DSP基础与应用系统设计．北京：北京航空航天大学出版社．2002
[53]TMS320DM642 Technical Overview. Texas Instrument Incorporated. 2002(9): 32-34
[54]Texas Instruments Incorporated. TMS320C6000 Code Composer Studio Getting Started Guide. 2003(8): 56-58