Predicting Travel Time of a Vehicle in Occluded Area with SUMO

The work aims to develop robust continuously vehicle-tracking algorithms with aerial images. It concentrates on predicting travel time of a vehicle travelling in occluded areas, such as tunnels and roads under bridges. Microscopic traffic simulation is used to estimate the movement of vehicles on roads, in the case that available information is limited or a large amount of factors are involved, such as intersections with traffic lights and accidents. According to preprocessed data from aerial images, roads, flows and traffic signals are defined and calibrated. Ultimately, the travel time exported from the simulation is regarded as the time to guide a glider to track the vehicle at the end of the occluded area. With outputs of the simulation, the queue length and speed entering a tunnel is compared and assessed by detecting and tracking data from aerial images, which are taken above Munich by 3K cameras on a glider of DLR.