Model-based vehicle tracking from image sequences with an application to road surveillance
Abstract
A model-based approach to vehicle tracking is proposed and applied to a highway traffic surveillance problem, which is motivated by current research in intelligent transpiration systems. Systems for traffic management and traveler information services require accurate and wide-area estimates of vehicle velocity and traffic spatial and temporal densities. A detection and tracking algorithm is developed that achieves good performance with complexity low enough for real-time implementation using inexpensive microprocessors. Detection thresholds are computed based on a statistical model for vehicle and background, and the theoretical detector performance is derived. The tracking algorithms filters position estimates from the detection algorithm using a simple vehicle dynamic model and the Kalman filter. Data association is accomplished with a nearest neighbor filter coupled with a lane-change handling logic.
- Publication:
-
Optical Engineering
- Pub Date:
- June 1996
- DOI:
- 10.1117/1.600747
- Bibcode:
- 1996OptEn..35.1723K