摘要
为实现基于道路视频摄像机视频处理的运动车辆精确测速,提出基于分道线长度、车道宽度和分道线消失点的道路云台摄像机参数自动标定方法,以对摄像机焦距、俯仰角、旋转角、离地高度等参数进行自动标定。实验结果表明,在应用于基于视频处理的运动车辆速度检测时,所提方法与雷达测速在平均误差以及误差标准差方面远小于我国公路测速允许的误差上限,验证了方法的准确性和有效性。
To realize the accurate speed measurement of moving vehicles by processing road video camera,an automatic calibration method of road pan-tilt-zoom camera parameters based on lane marker length,lane width and lane marker vanishing point was proposed,which automatically calibrated camera focal length,tilt angle,pitch angle and ground height.Results of experiment in its application to vide-based moving vehicle speed detection demonstrates that its achieved average error relative to the ground truth obtained by radar and the standard deviation of the error are far below the upper bound of the nation’s speed error limit,verifying the accuracy and effectiveness of the proposed method.
作者
吴建平
陈珂
刘业
WU Jian-ping;CHEN Ke;LIU Ye(College of Computer Engineering,Suzhou Vocational University,Suzhou 215104,China)
出处
《计算机工程与设计》
北大核心
2021年第7期2069-2076,共8页
Computer Engineering and Design
基金
国家自然科学基金项目(61672372)
苏州市工业应用基金项目(SG1819)。
关键词
道路摄像机参数标定
云台摄像机参数标定
智能交通系统
道路车辆速度检测
视频检测
road camera calibration
pan-tilt-zoom camera calibration
intelligent transportation system
road vehicle speed detection
camera calibration
