Modeling Pedestrian Behavior and Detecting Event Anomalies using a Seasonal Kalman Filter
Abstract
We present a seasonal state-space model using Kalman recursions to learn and predict structured behavior patterns. The model is employed to detect events using the learned expectations of typical scene activity. We demonstrate the approach for modeling the expected number of pedestrians in a scene over hour-long periods (over multiple days) and for detecting event anomalies. The framework provides a single long-term model by exploiting the natural seasonal trends in daily human activity.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2005
- Accession Number
- AD1001143
Entities
People
- James W. Davis
- Mark A. Keck
Organizations
- Ohio State University