Sensor Networks for Transient Public Events

Substantial progress in signal and sensed data analysis has been made over the past two decades, albeit with a few operational restrictions constraining user expectations of the technology. These capabilities are much more powerful when placed within a network of sensors that are able to extract data at multiple locations and inter-relate that data in meaningful ways to support a multitude of applications. This paper provides a framework for handling data in sensor networks, and shows how this framework drives the design of a system to serve a real-world commercial application, the measurement of transient public events. While the approach described herein is inherently general, the sensors used in live deployments of these networks by Micro Target Media Interactive, are video cameras coupled with processors leveraging computer vision technology which automatically interprets video data.