Intelligent Engineering Systems through Artificial Neural Networks, Volume 16
99 Anomaly Detection for Surveillance Applications
Download citation file:
- Ris (Zotero)
- Reference Manager
The paper describes an automatic surveillance system. Special emphasis is placed on the description of the anomaly detection method developed that allows detecting novel, never previously encountered anomalies in people's behavior. The method employs Support Vector Machines to learn the model of each person's normal behavior from previously collected data. Results are described in the context of a facility surveillance application.