The number of cars on the roads has increased in the last several years. There’re many traffic jams in the cities, which make citizens’ life very uncomfortable. One of the reasons of traffic jams is low channel capacity of crossroads. It can be raised by the reconstruction of the crossroads. The reconstruction is based on the statistical information about the crossroad usage. It is needed to know what types of cars use the crossroad, how many cars move in different directions, their speed, etc.
One of the ways to gather the statistical information is to ask people about the directions they drive their cars, but its accuracy is very low and the price is high. The other way is to use computer vision. The road CV systems which are used nowadays aren’t eligible for this task, because they are set only on few lanes, while complex analysis of the whole traffic is needed here.
In this work we propose to use video stream recorded from above. We defined requirements, proposed architecture for the system. To choose which tools to use to implement the system, we compared NI LabVIEW IMAQ functions to those of OpenCV.
| Файл тезисов: | text.pdf |
| Файл презентации: | pres.pdf |