基于路网卡口系统车牌识别数据的机动车出行轨迹提取方法  

The Extraction Method of Travelling Path Based on Identification Data of the License Plate from Bayonet System

作　　者：胡正华 刘良旭[1] 王快[2] 

机构地区：[1]宁波工程学院,电子与信息工程学院,浙江宁波 [2]重庆市勘测院,重庆

出　　处：《测绘科学技术》2018年第4期356-362,共7页Geomatics Science and Technology

摘　　要：机动车的出行轨迹因其隐含了大量车辆出行的状态信息,已经成为研究城市交通流时空分布特性的重要数据来源。如何提取车辆的历史行车路径已经成为交通领域的研究热点。目前,有不少研究学者利用路网卡口系统的车牌识别数据来提取车辆的出行轨迹。本文在现有方法的基础上,提出了一种当识别数据中存在较多的粗差,且路网数据和卡口点位不能完全匹配的情况下,提取车辆出行轨迹的方法。实验证明,该方法不仅能够全面系统地再现复杂的交通运行场景,为基础OD矩阵的调查和更新工作提供思路和有效的技术手段,还能为交通部门制定相关的决策、法规提供更为可靠的数据支撑。As it implies a large amount of information about state of the traffic, the trajectories of the vehicles have become an important data source for studying the spatial-temporal distribution characteristics of traffic in urban areas. How to extract the historical route of vehicles has become a research hotspot in the field of transportation. Currently, many researchers use identification data of the license plate from bayonet system to extract the trajectories. Based on the existing methods, this paper proposed a strategy for extracting trajectory from identification data with poor data quality and even when the road network cannot completely match the bayonet station. Experiments show that this method cannot only reproduce complex operation scenarios of the traffic comprehensively and systematically, provide ideas and effective technical means for investigation and update of the basic OD matrix, but also provide more reliable data support for the transportation department to make relevant decisions and regulations.

关 键 词：轨迹提取 数据挖掘 智能交通系统 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]