TY - JOUR TI - A spatial-temporal-map-based traffic video analytic model for large-scale cloud-based deployment DO - https://doi.org/doi:10.7282/t3-pap7-ty40 PY - 2020 AB - The CCTV (Closed Circuit Television) traffic surveillance systems are one critical part of traffic operations and management at state and local DOTs (Departments of Transportation). In this thesis, a multi-lane traffic detection model based on the spatial-temporal map (STMap) is proposed, which is a few longitudinal scanlines instead of the entire video frame to detect vehicle trajectories. The proposed model is built based on a prior STMap-based aerial video analytic method but with significant improvement addressing issues with roadside CCTV traffic cameras. A motion-flow-based direction determination method, a bi-section occlusion detection and splitting algorithm, and the lane-changing track method are proposed for determining the directions, splitting vehicles from occlusions, and identifying lane-changes. The model evaluation is conducted by using videos from multiple cameras from the NJDOT (New Jersey DOT)’s 511 traffic video surveillance system. The results show promising performance in both accuracy and computational efficiency for potential large-scale cloud deployment. KW - Traffic cameras KW - Civil and Environmental Engineering LA - English ER -