基于修正ASM的驾驶员警惕性识别方法研究

针对驾驶员警惕性研究中分析注意力程度的重要性,提出基于修正主动形状模型（ASM）的驾驶员警惕性识别方法。首先建立包含26个特征点的人脸ASM,其次结合面部结构约束构建了双眼平均合成精确滤波器（ASEF）,并通过旋转进一步增强鲁棒性,然后用改进ASEF修正人脸ASM;采用左右瞳孔和鼻子特征点建立三角形视线模型,并分析驾驶员注意力程度,利用左右眼角特征点距离对眼睛闭合程度进行归一化,最后利用支持向量机（SVM）分类得到警惕性程度。利用Visual Studio 2017平台进行实验,结果显示,改进ASEF滤波器的准确率达到95.16%,SVM对警惕性程度的分类准确率达到93.8%,每帧平均耗时49.13 ms,表明提出的方法能够有效地识别驾驶员的注意力程度以及警惕性程度。     

驾驶员疲劳检测系统研究综述

实时检测驾驶员的状态,可以避免大量交通事故的发生.通过机器视觉技术可以识别出驾驶员眼睛张开闭合状态,判断驾驶员是否疲劳.基于视频图像分析的驾驶员疲劳检测方法成为当前研究的一个重要方向.本文对国内外驾驶员疲劳检测方法及发展趋势进行了综述.     

基于计算机视觉的疲劳及注意力检测算法

针对现有疲劳驾驶检测算法的准确率底、实用性差以及不能实时检测的问题,论文采用计算机视觉的方法首先利用Dlib提取眼部、嘴部、下巴周围的特征点的坐标,通过计算眼睛纵横比来实现瞌睡检测,并且类比眼睛纵横比提出一种用于哈欠检测的新方法—嘴部纵横比检测法,为了进一步判断驾驶员状态,论文进一步对驾驶员的注意力进行判断,一旦出现不安全行为,立马发出预警进行提醒,从而实现驾驶员疲劳及注意力检测.实验证明,该方法在疲劳驾驶检测的准确率、实时性等方面取得明显的性能提升.     

基于 HOG 特征提取与 SVM 驾驶员注意力分散行为检测方法研究

驾驶员注意力分散是导致交通事故的主要原因,该文以驾驶员注意力分散行为图像为分类 目标,提出了一种基于方向梯度直方图(Histogram of Oriented Gradient,HOG)与支持向量机(Support Vector Machine,SVM)的行为检测方法。首先,获取图像中的感兴趣区域,并对图像进行增强、去噪及归一化处理;然后,提取图像 HOG 特征,进而采用交叉验证法对 SVM 分类器中的参数进行优化;最后,对视频图像中驾驶员的不同行为进行分类识别。实验中,通过与传统 SVM 算法以及基于局部二值模式的 SVM 算法进行对比,验证了所提方法具有更好的识别准确率。     

基于贝叶斯网络的疲劳度及注意力检测

为提高基于单一特征检测算法的准确率和可靠性，提出基于贝叶斯网络融合多个特征参数的检测算法。定位眼睛和嘴巴，利用两眼和嘴巴组成的三角形建立头部旋转模型，提取各特征的参数并用贝叶斯网络进行融合，用来判断驾驶员的驾驶状态，当出现非正常驾驶状态时给以警告。实验结果表明，该算法对于检测出驾驶员的疲劳度和注意力分散状态有较高的准确性。     

基于计算机视觉的驾驶员低头行为检测

在驾驶过程中使用手机会引起驾驶员的注意力分散,为了对这种行为进行监督和提醒(在公共交通中检测更有意义),提出了一种基于脸部特征提取的驾驶员低头行为的检测方法。该方法使用主动型状模型(Active Shape Model,ASM)算法得到脸部特征点,在此基础上通过脸部特征点的位置信息计算出头部姿势描述信息,最后通过SVM将上述信息分类进而得出头部姿势,其可用于判断驾驶员是否在驾驶过程中低头看手机行为,该方法能够有效检测出驾驶员在驾驶过程中低头使用手机的行为。实验结果表明,该方法的平均检出率在94%以上。     

驾驶员眼部状态检测研究

对拍摄得到的驾驶员视频帧图像,使用复合肤色模型检测人脸;通过自适应边缘检测、图像增强等方法处理得到特征图像,经特征区域筛选,依据人脸先验知识匹配得到最佳人眼对;提取眼部特征向量,结合LVQ神经网络进行模式识别检测眼部状态,为判断驾驶员是否处于疲劳状态提供判据。     

瞳孔亚像素边缘检测与中心定位

为提高视线跟踪系统的分辨率和精确度,提出了一种亚像素精度的边缘检测与中心定位算法。利用设计的一维边缘检测算子对瞳孔像素级边缘点进行提取,然后利用边缘点小邻域内梯度方向上图像信息,进行亚像素边缘位置检测,最后经椭圆拟合确定瞳孔中心的精确位置及瞳孔形状参数。仿真实验验证了算法的有效性,提出的算法已被成功应用于实时视线跟踪系统中。     

改进的遮挡条件下瞳孔检测方法

受到眼睑、睫毛、反光等干扰,提取的瞳孔往往无法近似成标准的椭圆,给瞳孔精确检测带来了难度。针对遮挡问题,提出了一种改进的遮挡瞳孔精确检测方法,该方法首先获取瞳孔图像的感兴趣区域,在对该区域进行阈值分割的基础上,通过滤波去噪和扫描线法去除光源反光的影响;根据沿遮挡瞳孔的分界线进行水平旋转,应用内接平行四边形法确定瞳孔的中心;根据圆形检测理论,提出一种中心修正算法,将求得的两个圆的交点作为眼睑与瞳孔的交点,计算出旋转角度,从而精化瞳孔中心的定位;最后利用五点法计算得到的瞳孔椭圆参数作为初始值,通过最小化拟合椭圆与边缘点之间的欧氏距离,使得该椭圆在非线性最小二乘意义下是最优的。实验结果表明,该方法具有良好的鲁棒性和准确性,在遮挡不超过瞳孔面积一半的情况下,能得到较为精确的瞳孔位置和边界。     

基于瞳孔-角膜反射技术的视线跟踪方法

视线跟踪是基于多通道的人机交互技术的重要研究内容．而基于瞳孔-角膜反射技术的视线方向是目前应用最广泛的视线跟踪技术之一。瞳孔-角膜反射技术的主要目的是提取人眼图像中瞳孔-角膜反射向量作为视线方向计算模型所需的视觉信息,通过搭建红外光源设备提取瞳孔-角膜反射向量构建基于瞳孔-角膜反射技术的视线跟踪系统,为面向人机交互的视线跟踪研究提供可行的低成本解决方案。     

Gaze direction estimation using support vector machine with active appearance model

In recent years, research on human-computer interaction is becoming popular, most of which uses body movements, gestures or eye gaze direction. Until now, gazing estimation is still an active research domain. We propose an efficient method to solve the problem of the eye gaze point. We first locate the eye region by modifying the characteristics of the Active Appearance Model (AAM). Then by employing the Support Vector Machine (SVM), we estimate the five gazing directions through classification. The original 68 facial feature points in AAM are modified into 36 eye feature points. According to the two-dimensional coordinates of feature points, we classify different directions of eye gazing. The modified 36 feature points describe the contour of eyes, iris size, iris location, and the position of pupils. In addition, the resolution of cameras does not affect our method to determine the direction of line of sight accurately. The final results show the independence of classifications, less classification errors, and more accurate estimation of the gazing directions.     

基于DM642的嵌入式疲劳驾驶监测系统的实现

针对基于DM642的嵌入式疲劳驾驶监测系统，讨论了系统的总体结构，提出了采用可控光源、奇偶差分帧和人眼跟踪的人眼检测方法。在与奇偶帧图像采集同步的内外圈光源的照射下，利用近轴光源的红眼效应，引起图像奇帧暗瞳孔、偶帧亮瞳孔，由奇偶差分帧快速实现人眼检测与跟踪。同时分析眼睛的特征参数，在一定时间内连续统计眼睛的闭合时间，计算眼睛累计闭合持续时间占某特定时间的百分率（PERCLOS）值来判断疲劳程度并报警。该系统疲劳监测准确率较高，可在汽车驾驶中进行实时疲劳驾驶监测。     

Windows视频驱动程序的技术剖析和比较

视频驱动程序是图像处理系统的基础环节,其效率、可靠性将直接影响图像处理系统的性能。文章论述了适合不同Windows操作系统的两种视频驱动程序的技术:适用于Win9x的基于虚拟环境的视频驱动程序技术和适用于Win2000/XP的基于WDM的视频驱动程序技术。并以开发BT878卡视频驱动程序为例,详细比较了两种视频驱动程序技术优缺点,基于WDM的视频驱动程序具有采集效率高、稳定性好、延时短和模块化程度高等许多优点,因此,基于WDM的视频驱动程序技术代表视频驱动程序技术发展的方向。     

基于视频识别驾驶疲劳的信息融合系统

本文提出并实现了基于实时视频检测驾驶员疲劳度的信息融合系统。使用红外摄像仪和高速DSP相结合,实时监控驾驶员的疲劳症状,提炼出特征信息,经FPGA系统加以信息融合技术处理后得到疲劳度,自动判断驾驶员是否进入疲劳状态,根据疲劳度的不同发出警告信号。经实际测试表明,使用信息融合技术有效地提高了系统判断驾驶疲劳状态的精确度和系统的鲁棒性,具有93%以上的正确率。     

Deus est machina [artificial intelligence]

What happens if the artificial intelligence community, in its quest to build intelligent systems, succeeds too well and creates an AI whose intelligence exceeds the threshold marked out by our own? Up to now, it is humans who develop the software and hardware and who drive all progress in capability. After crossing the threshold, however, the AI itself will rapidly augment its own capabilities. What's the intuition here? Although we use technology to help us conceptualize, design, and build today's computers and software there's no doubt that we remain in the driver's seat. But imagine the software design process reaching a level of complexity at which human designers exert only executive oversight.     

WebTracer: A new web usability evaluation environment using gazing point information

WebTracer is a new usability evaluation environment that supports recording, replaying, and analysis of a gazing point and operation while a user is browsing a website. WebTracer can record a user’s gazing point and operation compactly.Results of an experimental evaluation showed that the size of the operation history recorded by WebTracer was from 1/10 to 1/20 of the size of data recorded by an MPEG-2 or MPEG-4 format. With its compact form, the results of usability testing with the gazing point can be efficiently shared. It is expected to easily share empirical data between researchers. Also, evaluators can easily send the testing results as a feedback to the developers over the Internet.Moreover, the results show a possibility that gazing points related to usability. For example, when the menu of a Web page is divided into two panes, gazing points move more quickly. It seems that WebTracer efficiently improves usability evaluation, with gazing point data helping to identify problems on Web pages.     

Intelligent process control via gaze detection technology

Gaze detection is to locate the position on a monitor screen where a user is looking. In our work, we implement it with a computer vision system setting a camera above a monitor, and a user moves (rotates and/or translates) his face to gaze at a different position on the monitor. In our case, the user is requested to keep the pupils of his/her eyes fixed when he/she gazes at a different position on the monitor screen; though we are working on to relax this restriction. Till now, we have tried different methods and among them, the proposed Two Neural Networks method shows the best results. For application of the gaze detection techniques to the user-interface of a control program, we supplement two additional techniques. One is to drag a mouse cursor by moving his/her face after initially placing the cursor by gazing, and another is to click the cursor by winking one eye. These are applied to user-interface of a process control program in three different ways. Firstly, even when both hands are busy in doing a work, the user can still control the monitor screen by gazing, dragging and winking. Secondly, if the user is absent and not in front of the monitor when important states occur, then the system can record and replay them later when the user returns. Finally, as the system knows where the user is looking, if an emergency occurs at a different point from where he/she is looking, then the system can display a warning signal. This paper also includes subjective and objective experimental results obtained from the tests by applying the underlined techniques to a process control for chemical vapor decomposition.     

基于激光测距仪的移动机器人避障新方法

针对向量场直方图（VFH）算法对阈值敏感的问题,基于激光测距仪提出一种通过自适应调节阈值求 可行方向的移动机器人实时避障算法．对于给定的一组阈值,用其中的每个阈值求可行方向,在获得的接近目标方 向的所有可行方向中,选取最大阈值所对应的方向作为机器人的参考行驶方向．根据当前行驶方向与参考行驶方 向之间的偏差和所选阈值的大小对线速度进行限制．在ASR 机器人上进行实验的结果表明,当最大运行速度设为 0.8 m/s 时,机器人能够以0.61 m/s 的平均速度安全平滑地通过障碍区．     

基于眼睛状态的驾驶员疲劳模糊识别研究

首先介绍了当前国内外各种驾驶员疲劳识别方法,并分析其优缺点后,通过对驾驶员眼睛状态的定位和跟踪,提取眼睑距离来得到眨眼的频率和间隔,提出了驾驶员疲劳的模糊识别算法判定驾驶员的疲劳程度,并利用摄像机实时拍摄驾驶员的脸部,试验表明,提出的模糊识别方法能快速、准确地判断甚至预测瞌睡的发生,对于一般驾驶时目标检测正确率可以达到95%,当驾驶员头部开始转动时,正确率会下降至87%,但仍在可以判别范围内,具有较快的响应速度和鲁棒性.     

Purposeful gazing and vergence control for active vision

The purposeful-gazing capability of active vision offers advantages to many manufacturing tasks. This paper discusses the problems associated with purposeful gazing and fixation of attention for active vision. In binocular active vision, gazing at a selected target refers to directing the visual axes to capture the target in the (appropriate part of) the visual field by both sensors (cameras), and holding gaze refers to directing the visual axes of the sensors so as to maintain the target or point of interest in the visual field of both sensors. This paper proposes solutions to the important problems involved in gaze stabilization by developing techniques for vergence error extraction, and vergence servo control. Vergence is the tilt movement process of two visual sensors (in a binocular system) in opposite directions to fixate at a selected point. Binocular gazing is realized by decreasing the disparity which represents the vergence error. In order to obtain the disparity for extraction of vergence error, a phase-based approach that robustly and efficiently estimates vergence disparity is developed. To control vergence, we present a dual sampling-rate approach for vision-sensor-based dynamic servo control.