论文部分内容阅读
目的研究一种基于图像分析技术的汽车驾驶员瞌睡报警装置,以实现当司机饮酒或长时间驾驶引起疲劳驾驶和瞌睡时,予以提醒报警,减少事故发生,并记录当时的汽车行驶参数与眼睛状态,可作为事故原因分析的依据。方法利用视频采集和数字信号处理(DSP)模块对视频流进行动态分析确定头部位置,通过图像分析方法跟踪眼睛的位置和判断眼睛的状态。结果利用硬件DSP和视频采集卡实现了检测驾驶员眼睑状态的功能,跟踪速度达25 f/s;经实际驾车测验,双眼在摄像机视野内的图像跟踪准确率达到96.9%,整个测量过程的平均跟踪准确率为95.7%。结论该装置提高了跟踪驾驶员眼睛位置的速度和准确率,并可识别其驾驶过程中的瞌睡状态。
OBJECTIVE To study a driver sleepiness alarm device based on image analysis technology in order to remind the police to reduce the number of accidents and record the driving parameters of the car and the eyes when the drivers drink or prolonged driving causes fatigue driving and drowsiness. , Can be used as the basis for accident analysis. Methods Video acquisition and digital signal processing (DSP) module are used to dynamically analyze the video stream to determine the position of the head, to track the position of the eye and to determine the state of the eye through image analysis. Results The detection of driver’s eyelid function was realized by using hardware DSP and video capture card with the tracking speed of 25 f / s. The actual driving test showed that the accuracy of both images tracking in the camera’s field of vision was 96.9%. The average of the whole measurement process Tracking accuracy of 95.7%. Conclusion The device increases the speed and accuracy of tracking the driver’s eye position and identifies drowsiness during driving.