基于时空车流灰度的城市交通热点分析

Urban Traffic Hotspots Analysis by Gray Image of Spatio-temporal Vehicle Flows

随着城市规模的扩张和汽车数量的增加,拥堵和雾霾等受交通影响的现象变得越来越严重,城市交通热点发现与分析是改善交通状况的关键技术之一。根据城市中典型浮动车(出租车)的时空轨迹数据,基于图像分析理论,提出城市交通热点的空间分布分析方法。首先,根据浮动车数据,形成车流轨迹的时空分布灰度图。其次,通过将浮动车的轨迹数据映射到高精度城市交通网格上,探索了轨迹数据和城市热点区域之间的关联关系。进一步地依据车流轨迹的时空分布,发现车流密度的极值。最后,结合影响区域的参数利用高斯曲面拟合对时空灰度图像进行了热点区域分析,获得城市交通热点的空间分布。将分析结果在地图上标定,准确地反映了城市交通热点区域的详细位置及空间分布,从而给出了一种有效、直观的城市交通热点分析方法。研究结果对于城市交通规划和交通实时信息发布具有实用价值。

With the rapid development of modern cities and the dramatic increment of the amount of cars, the phenomena of congestions and atmospheric hazes due to traffics become much worse. Urban traffic hotspots detection and analysis is one of the key technologies to study traffic situations. In this paper, we analyze the spatio-temporal trajectory data of urban floating cars and apply the method of image analysis to investigate the spatial distribution patterns of these urban hotspots. First, we build up the gray images based on the spatio-temporal distribution of full data sets on floating cars. Then, we map the trajectory data to urban traffic grids with high degree of accuracy, in order to find out the correlation between the trajectories and urban hotspots. Furthermore, the extremum points of vehicle density are detected based on the spatio-temporal distribution patterns of the trajectories. Finally, we discover the most frequented locations by Gauss curve fitting on spatial-temporal gray image with the range of influence of traffic hotspots, and detect the spatial distribution of urban traffic hotspots. The proposed method is validated and the results indicated that the method is effective and can be used to visualize urban traffic hotspots.