This thesis introduces a video-analysis based traffic monitoring system. With the widespread development of information technology in the recent decades, more and more real life processes are automatically analysed for purposes of analytics and optimization. Traffic analytics is such an area where next to the so far conventional methods (optical tripwire counters, inductive loops) video analytics is gathering more and more momentum.
With the increasing processing power and decreasing cost of information technology equipment, video analytics is getting to be an ever increasing cost effective solution. For the installation, there is no construction is required on the road surface, the only hardware items to be installed are the cameras and their supporting network infrastructure. The functionality is defined by the software so later modification or extension of it does not require costly on-site operations.
Fusing the events and state information of a traffic intersection into a bigger system helps the operators to understand the real-time situation and enables them to respond in a timely manner. Time series of traffic data gathered through an extended period can provide insight into the utilization of certain road sections and can support long-term traffic routing decisions.
The thesis shows the design and implementation of such a system from the analytics to the map based display of traffic states and events.