This process had several components: 1. Track each pedestrian throughout scene using the Kalman filter estimates. 2. Record the position and velocity state. 3. Develop a position and velocity path characteristic for each pedestrian. This was done using the Kalman filter prediction of future state. 4. Set a “warning signal” under the following conditions: a. Pedestrian enters near a “secure area” (a gate, expensive art display, podium, etc…). b. Pedestrian moves above a walking speed . c. Pedestrian loiters in the area for a long time. d. Pedestrian falls down.