Model-based vision for automatic alarm interpretation

In a previous paper T.J. Ellis et al. (1989) described the development of a knowledge-based vision recognition system for automating the interpretation of alarm events resulting from a perimeter intrusion detection system (PIDS). Measurements extracted over a sequence of digitized images are analyzed to identify the cause of alarm. Models are maintained for both alarm causes and the scene and the measurements are matched with the models to derive an appropriate classification of the event. The authors record progress on the further development of the system and present the results of applying it to a number of real alarms. The system is shown to behave robustly, correctly classifying genuine alarm events (i.e., human intruders) and providing statistics of false alarm events