Video Surveillance as a Means of Detecting Fires
Current computational technologies open new possibilities in many traditional applied areas, such as fire safety. The problem of fire detection has been a long standing one, and may be approached in many different ways. Video observations are a modern approach, which is closely related to Artificial Intelligence (i.e. reminiscent of fire of fire detection by humans). Some work has been done already in this direction, however, developments are still rudimentary and the underlying technologies need great improvement.
The nature of the project will be to generate realistic fire scenarios (“fire signatures”), experimentally or computationally, and then develop an Artificial Intelligence methodology that will allow identifying fire ignition events.
Modern technologies of video image processing will be used. In particular, video-based scenario recognition techniques from the SAVASA (funded by EU FP7 Security Programme, 2011-2014) project will be extended and then used.
Personnel Involved
First Supervisor: Liu, J Dr
Second Supervisor: Novozhilov, Prof V, Faculty of Art, Design & the Built Environment
Third Supervisor: Wang, H Dr
Collaboration: This project does not involve collaboration with another establishment