论文研究-驾驶员疲劳检测技术研究综述.pdf

所需积分/C币:50 2019-07-22 20:25:48 228KB .PDF
收藏 收藏 3
举报

针对近几十年来驾驶员疲劳检测的发展状况,对驾驶员疲劳检测技术进行综述。通过将疲劳检测技术分为基于驾驶员特征以及基于车辆行为的疲劳检测技术,详细阐述了相关研究成果。基于驾驶员特征的疲劳检测通过提取包含疲劳信息的特征进行检测,具体分为生理参数特征以及视觉特征两方面,而基于车辆行为的疲劳检测是根据车辆行驶情况间接判断驾驶员是否有疲劳。介绍了几个典型疲劳检测系统,最后探讨了疲劳检测的技术难点以及发展趋势。
第5期 杨海燕,等:驾驶员疲势裣测技术研究综述 1623· 这些方法的主要问题在于对于不同车型、不同光线、不同展,需要加强对疲劳驾驶视频图像数据库的建设,统一系统评 的道路及路面状况下,婁达到比较好的效果具有很大难度。 测方法和规范。 3典型检测系统介绍 5结束语 目前疲劳检测系统比较多,效果比较好的有以下几种: 本文综述了驾驶员疲劳检测技术目前主要研究方向。尽 a)澳大利亚国际大学开发的DAs在商业上获得使用。管在过去的二十多年中,疲劳检测研究取得了突破性进展,但 使用安装在汽车仪表板上的人眼跟踪系统监控司杋,能监控驾目前仍然存在许多冇待解决的冋题。由于驾驶环境差异以及 驶员表现,利用方向盘的握力反馈,同时将路面跟踪偏差反馈从驾驶舒适度的角度来说,目前的基于机器视觉疲劳检测方法 给驾驶员。路面跟踪设备目的是检测车辆突然擦过路面标志比基干驾驶生理参数以及车辆行为的方法更受关注。在今后 或者道路边缘等违规行为,该系统在检测到异常时给出警告,的研究中,无论是对基丁驾驶员自身特征的这类疲劳检测方法 并有方向鳌的制动功能。同时釆用的座椅振动报警与道路侧还是对丁基丁车辆行为的疲劳检测方法,疲劳特征的挖掘、提 而偏离的程度有关,以鼓励他们纠正航线偏离。 取以及与统计分类算法的结合研究将成为疲劳检测研究的重 1)欢盟于204年完成 AWAKE T程。采用的特征状要研究内容。同时要将疲劳检测从模拟驾驶走向真实环境的 态包括眼睑运动、握力改变及路面跟踪,并使用刹车和方向盘驾驶,克服驾驶员个体差异以及驾驶环境,提高检测的实时性 位置等制动行为,把这些方法结合起来抵制交通风险。 以及准确性,提高疲劳检测的准确性、鲁棒性以及速度,使得疲 c) Carnegie Mellon大学开发的 Copilot监控系统。通过劳检测性能得到大幅提高并能成功应用到实际驾驶中去。笔 Perclose测量眼睑运动,监控系统小且好用,提供一种有效的者相信,在研究者的努力和实际应用的需求的带动下疲劳检 研究工具。 测技术必将取得更大的发展、发挥更大的作用。 d) Seeing machine研究组开发的 FacelAB系统监控驾参考文献: 驶员行为,能检测疲劳与精力分散等情况。用一对视频相忛获 [1] WANG Qiung, YANG Jing-yu, REN Ming-wu, et al. Driver fatigue de- 得视频图像,从左到右匹配得出每个特征的三维位置。采用最 tection: a survey C//Proc of the 6th World Congress on Intelligent 小二乘优化法定位头部三维位置, aceLA软件并行处理眼睛 Control and automation 2006.21-23 凝视数据,定位虹膜中心,根据眼睛凝视向量确定哏睛凝视方[2]I.AI.SKI, CRAIG A. A critical review of the psychophysiology of 向,计算眼睛张开以及眨眼频率,监控眼睑。 Facula在模拟 driver faligue[ J. Biological Psychology, 2001, 55(3): 173-194 驾驶中被证明非常有效,且在低光线、头部大范围运动以及视3」 LAL S KL, CRAIG A. Electroence phalography activity associated 觉方向跟踪上有很好的效果,即使司机带太阳镜也能检测出。 with driver fatigue: implications for a fatigue countermeasure device 日前 FacelaB已经有产品43.以及工具箱下载并提供疲劳检 [J. Journal of Psychophysiology 2001, 15(3): 183-189 测程序的外部接口 [4 LAL S K L, CRAIG A, BOORD P, et al. Development of an algorithm 可以看出,系统利用各种视觉特征以及车辆的行为表现等 for an EEG-based driver fatigue countermeasure[J]. Journal o Safety Research, 2003, 34(3): 321-328 综合特征去检测是否有疲劳产生,同时有的还采用某些报警方 式提醒驾驶员注意行车安全。 [5 FISCH B J. Spehlmalln's EEG primer[ M]. 2Id ed. amslerdlaIl Elsevier Science Ltd..1991 4存在的问题及发展趋势 [6 GU Hai-song, JI Qiang. An automated face reader for fatigue detection [C//Pron of the 6th IEEF International Conference on Automatic 由于驾驶员个体差异以及光线、路面等驾驶环境差异大 Face and Gesture Recognition. 2004: 111-116 日前的痘劳检测算法基本都是基丁模拟驾驶环境,下一步要向17 JAP B T, LAL SKL, FISCHER P, ct al. Using EEC spectral compo 真实驾驶环境上发展,使疲劳检测技术广泛应用到商业领域 nents to assess algorithms for detecting fatigue[J]. Expert Systems with Applications, 2009, 36(2): 2352-2359 减少疲劳驾驶而造成的交通事故 [8 MAO Ming-wang, DU Li-ping. Research on drive fatigue detection 不论是从驾驶员自身特征还是从车辆行为方面看,可以直 using wavelet transform[C1//Proc of IEEE International Conference 接获得的特征类型有限,而且直接提取的表面特征效据多且有 on Vehicl lar F. lertronirs and Safety. 2007: 1-40 冗余。因此一方面要对疲劳特征进行挖掘,用先进的信号处理[91 shen Kai-4m, LI Xiau-piug, ONG ChUnIE-iinl,eta.EC1ae 方法提取最能表征疲劳的特征参数,另一方面要采用信号融合 mental fatigue me as urement using multi-class support vector machines 处理方法,将多个疲劳特征参数结合起来去对驾驶员疲劳状况 with confidence estimate[J. Clinical Neurophysiology, 2008, 119 进行检测,克服空间、光照等影响,提高检测算法的实时性、准 (7):1524-1533 确度 [10 YEO M V M, LI Xi shen Kai al Can SVM b 驾驶员疲劳检测系统应该具有疲劳决僉功能。由于不同 for automatic EEG detection of drowsiness during car driving I 人有不同的疲劳表现特征,疲劳检测系统应该具有智能,有自 Safety Science, 2009, 47(1): 1 15-124 学习、推理功能。在驾车初期,系统根据获取驾驶员的相关薮 [111 ZHANG Chong, ZHENG Chong-xun, YU Xiao-lin. Automatic recogni- tion of cognitive fatigue from physiological indices by using wavelet 据对系统进行训练,能得出驾驶员的疲劳特征选择最适合他的 packet trans form and kernel learning algorithms[ J]. Expert Systems 检测方法,能适应每个驾驶员。 th Applications,2009,36(3):4664-4671 目前公用疲劳驾驶数据集比较缺乏,检测系统性能泙价方 [12 LAL S K L, CRAIG A. Repruducibilily uf the spectral components of 法不统一,难以对各种疲劳检测技术的性能进行定量比较。为 the electroencephalogram during driver fatigue[J. International 了定量、客观地比较各种检测技术,推动疲劳检测方法的发 Journal of Psychophysiology, 2005, 55(2): 137-143 1624 计算机应用研究 第27卷 [13 OHSUGA M, KAMAKURA Y INOUE Y, et al. Classification of blink tion[C]//Proc of the 5 th International Conference on Transportation aveforms toward the assessment of driver 's arousal levels, an eoc stems.2002:168-173. approach and the correlation with physiological measures[ C1// Proc of [29] ITO T, MITA S KOZUKA K, el ul. Driver blink Measurement by the the 7th International Conference on Engineering Psychology and Cog motion picture processing and its application to drowsiness detection nitive Ergonomics. Berlin: Springer, 2007: 787-795 C//Proc of International Conference on Intelligent Transportation 14 NOGUCHI Y, NOPSUWANCHAI R, OHSUGA M, et al. Classification Systems.2003:168-173 of blink waveforms towards the assessment of driver s arousal level. an I 30 CHU Jiang-wei, JIN Li-sheng, TONG Bin-liang, et al. A monitoring apprach for HMM based classifica lion fru blinking videv sequence method of driver mouth behavior based on machine vision[C1//Proc [CJ//Proc of the 7 th International Conference on Engineering f IEFE Intelligent Vehicles Symposium 2004: 351-356 Psychology and Cognitive Ergonomics. Berlin: Springer, 2007: 779 [31] KITHIL P W JONES R D, JONE M. Development of driver alertness 786 detection systems using overhead capacitive sensor arrey. SAE Techni [15 ROGADO E, GARCIA J L, BAREA R, el ul. Driver faligue deletIon cal Paper Series 982292[R.[S advanced Safety Concepts Inc system[C//Proc of IEEE International Conference on Robotics and 1998. Biomimetics. 2009:1105-1110 I 32 SMITH P, SHAH M, LOBO D V. Monitoring head/eve motion for [16 JEONG I C, LEE D H, PARK S W, et al. Automobile driver's stress driver alertness with one camera[C]// Proc of the 15th International index provision system that utilizes electrocardiogram C J//Proc of Conference on pattern recognition 2000. 636-642 IEEE Intelligent Vehicles Symposium. Istanbul:s n., 2007: 652 I 33 PILUTTI T, ULSOY A G. On-line identification of driver state for 656 lane-keeping taske[ C//Proe of A merican Control Conference. 1995 17] TOSHIYU KI M, MASAAKI M. ECG monitoring of a car driver using 678-681 capacitively-coupled electrodes[ C]//Proc of the 30th Annual Inter- [34] PILUTTI T, ULSOY A G. Itleutificatiun of driver stale fur lane-keeping tional Conference on Engineering in Medicine and Biology Society tasks[ J]. IEEE Trans on Systems, Man, and Cybernetics: Part 2008:13151318 18 LOCCHI C, ROVETTA A, FANFULLA F. Physiological parameters [35 BERTOZZI M, BROGGI A Real-time lane and obstacle detection variation during driving simulations C1//Proc of IEEE/ASME Inter- the GOLd system[ C//Proc of IEEE Irklligent Vehicles nalional Conference un Advanced Intelligenl Mec haLrunics. 2007: 1-6 Symposium. 1996: 213-218 L 19 ANON. Perclos and eyetracking: challenge and opportunity [R. Bed [36 BERTOZZI M, BROGGI A GOLD: a payAN e sereo vision system for generic obstacle and lane dect ford Applied Scienee Laboratories, 1999 Trans on [20 DINGES D F, MALLIS M, MAISLIN G, et al. Evaluation of techniques Image Processing, 1998, 7 [37]WLESOMA W S, KODAGONA YINS, BAXSUNYX A P Road for ocular ineasueinent as au index of fatigue uud the basis for alert ness manage ment J. Safety Science, 2009, 47(1): 115-124 boundary detection and trackirkk uslog lad& seisin IEEE Trans on Robotics and Atomation, 2004 56-464. [21 JI Qiang, YANG Xiao-jie Real-time eye, gaze and face pose tracking for monitoring driver vigilance[ J]. Real-time Imaging Archive [38]MAO Ming-wang pu Li-ping n driver fatigue detection using wavelet trausormaiun beseech on driver fatigue detection 2002,8(5):357-377 LC interence on yehiel lar electronics [22 WIERWILLE WW, ELLSWORTH L A, WREGGIT SS, et al.Re search on vchicle- based driver status/performance monitoring: [39] HARIKEX Z F On using Hough transform for driving development, validation, and refinement of algorithms for detection of tio cl//Proc of the 4th International Conferenc driver drowsiness, DOT HS &8247[R.[S1.]: National Highway eationalomputer Communication and Proces sing. 2008: 9 Traffic Safety Administration, 1994 [23 DINGES D F, GRACE R. PERCLOS: a valid psychophys YOUMIN K, HAHN M. Detecting driver fatigue based on the measure of alertness as assessed by psychomotor vigilante[ response and the front view environment of auto mobile[C]// Washington DC: US Department of Transportation, Federal\ Higway Proc of the 2nd International Symposium on Universal Communica Adninistraliun 1998 tion.2008:237-240 1 24 FABIO L C. Class I infrared eye blinking detec nso 141 FLETCHER L, APOSTOLOFF N. PETERSSON L, ct al. Vision in and ActuatorsA:Physical,2008,148(2)/439 out of vehicles[ J]. IEEE Trans on Intelligent Transportation Sys [25] HEITMANN A, GUTTKUHN R Technologies monIto tems,2003,18(3):12-17 prevention of driver fatigue[ C]/ re of e t InNgalroma Driv ing I 42 BOVERIS S Driver fatigue manitoring technologies and future ideas Symposium on Human Factors in DYer Training and [C]//Proc of AWAKE Road Safety Workshop. 2004 Vehicle Design. 2001: 8186 [43 ONKEN R. DAISY: an adaptive knawledge-hased driver monitoring [26] BOVERIE S, LEQKLENM, VIRIA gent systems for vide and warning system[ C]//Proc of International Conference on Vehicle monitoring ProM International Congress and Navigation and Inform ation Systems. 1994: 3-10 ols and Vehicle Navigation Systems [44 LONGHURST G. Understanding driver visual hehaviour[R. Brad 199g don, Australia: Seeing Machine Ply Linited, 2002 [27] Rng-n,UoN-you,ShiShu-ming,etal.Amonitoring[45]Facelab[Eb/Ol].[2009].HttP://www.seeingmachines.com/ behavior based on machine vision[C]//P roc product/ facelab/ UNEN lulellMenl Vehicles Symposium. 2003: 110-113 46] MACHOVER A, ENCARNAAO M. Tools and products, computer [28NITAT,RNWKA S. NAKANO K, et al. Driver blink measurement by graphics and applications J. IEEE Publication, 2006, 26(1) d its application to drowsiness detec

...展开详情
试读 4P 论文研究-驾驶员疲劳检测技术研究综述.pdf
立即下载 低至0.43元/次 身份认证VIP会员低至7折
抢沙发
一个资源只可评论一次,评论内容不能少于5个字
  • 至尊王者

    成功上传501个资源即可获取
关注 私信 TA的资源
上传资源赚积分,得勋章
最新推荐
论文研究-驾驶员疲劳检测技术研究综述.pdf 50积分/C币 立即下载
1/4
论文研究-驾驶员疲劳检测技术研究综述.pdf第1页

试读结束, 可继续读1页

50积分/C币 立即下载 >