基于车辆传感网的交通路网自动提取方法研究

Automatic Extraction of Traffic Network Based on Ubiquitous Interconnected Vehicular Sensor Networks

为适应交通路网规划和智能驾驶等应用的需要,基于泛在互联车辆传感网对交通路网及其边界进行自动提取。车辆传感网中的运动车辆对其自身位置进行实时感知,并将感知数据经泛在互联网络汇聚至应用服务器,应用服务器根据同一时刻的运动车辆的位置信息构建一幅幅实时分布场景,并将适当时间间隔的多幅实时分布场景迭加成一幅增强分布场景,并在增强分布场景中沿不同的方向搜寻路网边界运动车辆节点,最后根据搜寻结果综合标定出交通路网边界节点,并据边界节点提取出交通路网边界。实例分析表明,该方法提取的交通路网及其边界基本准确,可为交通路网规划和智能驾驶等应用提供支撑。

In order to meet the needs for traffic network planning and intelligent driving applications,the traffic network and its boundaries are automatically extracted via the ubiquitous interconnected vehicle sensor network.The moving vehicles in the vehicular sensor networks perceive their own position in real time,and converge the perceptual data to the application server through the ubiquitous interconnected network,and the application server forms a real-time distribution scene in accordance with the location information of the moving vehicles at a certain time,and superposes a number of real-time scenes separated by the appropriate interval into one overlapping distribution scene.The boundary vehicle nodes are searched in the different directions in the overlapping distribution scene.Finally,the traffic network boundary nodes are calibrated synthetically according to the search results of boundary vehicle nodes,and the traffic network boundary is thus extracted according to the boundary nodes.The example analysis indicates that the extracted road network and its boundary by this method are basically accurate,so the method can provide support to the applications of traffic network planning and intelligent driving.