About the Project
This project is about developing a new methodology for managing intelligent systems of distributed synchronous cameras. Multi-camera systems are increasingly used to identify emerging risks in large buildings and areas where many people walk and interact through
successions of corridors and open spaces. [1,2] Their applications range from monitoring patient well-being in hospitals to tracking antisocial behaviour in retail centres and detecting terrorist activity at airports. Systems of pan-zoom-tilt cameras used in such applications are very complex and notoriously tedious and error-prone to monitor and continually adjust by human security agents. We propose a PhD project that will develop a methodology to automate the evaluation of the activity of individuals and groups using complex autonomous multi-camera monitoring systems. The PhD candidate will develop:
1. Distributed algorithms for monitoring individual and group activities and event detection from multi-camera video sequences. This part of the project will extend existing algorithms for the identification of human activity  from single-camera video sequences devised in a previous project led by AB. Multi-camera systems will enable better capabilities such as those provided by 3D modelling of group activities 4 and the tracking of unfolding events through complex networks of cameras. Dynamic modelling on graphs will be used to model changing patterns in movement.
2. Model-driven engineering techniques for the dynamic reconfiguration of camera parameters such as pan-tilt angles and zooming, to improve the scene observation and to track complex events involving multiple individuals. Building on recent research led by RC, [5,6] this project component will use runtime stochastic modelling and verification to continually assess the risk situation and adjust the camera configurations accordingly. This will allow multi-camera systems to follow unfolding events and to react to adverse changes such as a camera being damaged accidentally or maliciously.
2 L. Bazzani et al., Joint Individual-Group Modeling for Tracking. IEEE Trans Pattern Analysis Mach Intell 37(4):746-759, 2015.
3 K. Stephens, A. G. Bors, Observing human activities using movement modelling, AVSS:44_1-44_6, 2015.
4 M. Grum, A. G. Bors, 3D modeling of multiple-object scenes from sets of images, Pattern Recognition 47:326-343, 2014.
5 R. Calinescu et al. Self-Adaptive Software with Decentralised Control Loops. FASE:235-251, 2015. 6 R. Calinescu et al., Formal Verification with Confidence Intervals to Establish Quality of Service Properties of Software Systems. IEEE Trans Reliability PP:1-19, 2015.
Why not add a message here
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.