Rvid System aims to integrate RfId and computer vision systems into a unique software platform. It is a video surveillance proof-of-concept application capable of tracking the position of the people inside a monitored scene through the fusion of computer vision and UWB RFid localization data. This allows, for example, to connect UWB active RFid tags with the image of the person who is carrying it. RVid also provides a face recognition system, integrated with RVid's computer vision module, for further analysis.
RVid is a proof-of-concept software designed to help a video surveillance operator in the task of monitoring multiple scenes. The software provides an inference engine that performs data fusion on multiple and heterogeneous sensors (computer vision systems and UWB RFid devices are here regarded as sensors) and guides operator's attention in order to selectively analyse the scene.
Furthermore, RVid assists operators with online analysis keeping trace of ancillar information like height and speed of a person, pictures of body and face and other less distinctive characteristics. A module that performs an identity verification process is also provided based on facial biometrics and all captured data is stored into a remote database for offline analysis.
RFid UWB integration with computer vision localization capabilities is the main innovation with respect to common digital video surveillance systems. This feature is particularly useful when both already known people (carrying a UWB tag) and unknown persons are allowed in a monitored scene. Robust and reliable information about identity and position of people, allow to split scenes into different areas defined through the concept of virtual borders, where a crossing is defined dynamically upon contextual information like risk and alarm levels and identity of the people and their behaviour.
Therefore Rvid's architecture makes it possible to infer context dependant information from the scene being monitored and to enhance the capabilities of the common video surveillance systems towards novel use cases, like for example monitoring of personnel safety inside risky areas.
For further information please contact the reference people.