Video surveillance systems rapidly develop as they need to provide security of goods and public and private places. These systems remained basic for a long time with sets of numerous cameras connected to several viewing screens in control rooms. For many years, supervising systems are being developed allowing security operators to focus their attention on zones of interest where alarms can be triggered according to the detected human activity such as intrusion. The corresponding captured images are generally analysed by the security operators. Although more and more video surveillance cameras are being installed, exploitation methods remain "manually" handled and the real time information extracted from the cameras are largely under exploited.
Using video surveillance, the VIDEO-ID project aims to achieve real time human activity detection including the prediction of suspect or abnormal activities. This project also aims to be capable to perform identification using face and iris recognition. Thanks to such identification, a detected person will be tracked throughout a network of distant cameras, allowing to draw a person's route and his destination. Without being systematic, a logic set of identification procedures is established: event's situation and abnormal behaviour, people face recognition and formal iris identification.
The actors of this project are all specialised in one of the steps required for the sequence of identification to be achieved.
- Define the concepts of identification use by video surveillance.
- Improve the intrinsic performances of each of the image analysis technologies.
- Build the automatic chain of operations required and evaluation of the systems.
- Widen the capabilities of the intelligent video surveillance systems in order to detect individuals with acceptable accuracy in all kinds of situation and allow for their identification.