In recent months, already poor security situation at public places in Pakistan is further deteriorating requiring immediate counter measures to be taken. In case of conventional camera networks used for security purposes, the requirement of human operators for monitoring limits wide spread deployments of these networks. This issue can be partially addressed by developing an automated distributed intelligent camera network that can extract valuable information from the captured data by performing local data processing in real time. It is well know that the large data volumes generated by camera nodes limit the performance of these networks.

A first step towards addressing the above mentioned issues is an event driven system reducing the data volumes generated by the camera nodes. Integrating pyro-electric infra-red (PIR) sensors with camera nodes will lead to an event driven node architecture. We will develop stochastic resource control mechanism for this integrated sensor node architecture achieving optimal utilization of limited resources available. Modular hardware test-bed will be developed for physical realization. Distributed algorithms will be developed for local data processing at the camera nodes for information extraction. This will provide an opportunity for real time activity monitoring to enhance the security of the public spaces.