视频方法鉴定车速的准确性研究

目的通过实验研究用视频方法鉴定车速结果的准确性及其影响因素。方法在选定的实验场地上,让实验车辆以一定速度匀速通过一段距离,使用高空、垂直、斜角和车载四个位置的摄像机同时录制,将车速计算结果进行比较。结果理论和实验数据分析表明,误差的大小主要取决于:车速和帧速率,车体上参考点和路上参考点的对齐效果,还有行车距离的大小和播放器屏幕的大小等。结论用视频方法进行车速鉴定,相对于传统的用力学公式进行计算,其准确性和可信度较好;一般条件下,误差可在5%~10%之间估算;夜晚条件下,可在车灯的轮廓线上选择车体上参考点,炫目情况除外。

Accuracy of Determining Vehicle Speed by the Involved Video

Objective To test the accuracy of determining vehicle speed by the involved video and explore the infl uencing factors.Methods At the selected field,the experimental vehicle was driven at a constant speed across a distance,with its moving course being recorded by four cameras that had been fi xed at different positions(1:a site at the high storey of a building beside the road that the vehicle rolled past;2:a side place vertical/slanting to the road;3:the vehicle-carried tachograph).Based on the frames and elapsed time from the recorded videos,the vehicle’s speed was respectively measured and compared.Results From the analysis of relevant theory and experimental data,the metrical errors mainly relate to the vehicle’s speed,video’s frame rate,aligning accuracy of the reference points selected on the vehicle against the ones on road,the vehicle-rolled distance,and the screen size of the video-playing machine.Conclusions In contrast with the traditional mechanical approach,vehicle speed determination by the involved video is of both better accuracy and reliability,with the average error between 5%and 10%.For the night circumstance,the reference points on the vehicle had better choose around the contour of the vehicle’s lights with the exception of dazzling.