Overview
Traditional monitoring techniques are consolidated and reliable, but can only measure a limited set of fundamental parameters. In the last ten years, traffic picture processing has spread rapidly because of its lower impact on infrastructure and users and new fields of applications it offers, such as access control, surveillance, adaptive management of signals, estimation of poor safety conditions. The technique is still young and can be significantly developed. The existing commercial devices have limits of applications due to implemented techniques, lower accuracy and reliability, their high costs and need of power supply. The application of motion-tracking techniques to traffic film is innovative compared to the technique of virtual coils, and can significantly reduce counting and classification mistakes, and increase the number of parameters that can be measured (e.g., analysis of trajectories). Their integration with wireless communication systems is also innovative and necessary.
The main objective of the project was to encourage the development of procedures, techniques and devices for measuring and transmitting traffic data which can support the management and operation of road network, providing innovative solutions in the field of measurement methods, devices and systems. The project tested motion-tracking techniques applied to traffic films in order to improve the robustness (reduction of the sensitivity to weather conditions like rain or fog) and accuracy in identifying traffic parameters of individual vehicles (e.g. identification, trajectories, etc.) and conditions (e.g. bottlenecks, poor safety, etc.).
Funding
Results
In the field of measurements based on picture elaboration, development of systems for artificial vision and 3D elaboration and their characterisation through definition of analytical models for a priori assessment of standard uncertainty, the project studied the effects of parameters affecting acquisition and quality of traffic films (vibration of props, fog, rain, etc.), implemented and tested motion tracking techniques which can identify vehicles in MPEG files (in order to measure traffic parameters such as counting, classification, speed, flow, etc.) and analysed their sensitiveness to influencing parameters.
The project also studied and tested sensing techniques for rain and fog, integrating sensitive devices into short-range wired and wireless interface modules. Finally, the project developed innovative methodologies for diagnosing short-range wireless transmitters, found new approaches for measuring power of short-range wireless communication systems and interferences in wireless communication channels.