Title: Multisensor data Fusion for Video Surveillance

Abstract:
The lecture aims to give an outline of multisensor automatic surveillance systems. In 
the first half, after an overview of their historical evolution, the general physical 
and logical architecture of automatic video surveillance systems will be presented and 
the main processing steps discussed. Examples and algorithms for motion detection, blob 
extraction and classification, tracking, object and event recognition will be presented
 together with real-world video sequences.

In the second half, the extension to the multisensor case will be addressed. The concepts 
of data and sensor fusion will be introduced, and their application to surveillance 
illustrated. A multisensor architecture that dynamically regulates the fusion process to 
take into account the performance of the sensors in detecting the targets will be discussed. 
A step further will also consider contextual information as an optimization means.

The remaining part of the lecture will cover the most interesting and active research 
areas with comments on future directions.