Real-time estimation of vehicle emission has practical significance in traffic management and information service. This paper proposes a video-based estimation method of vehicle emission where video installed on the roadside is used to track vehicle trajectories in real time and vehicle specific power (VSP) is applied to estimate vehicle emission. First, vehicle trajectories in a specified road section are obtained by a video-based algorithm. Three key steps of the algorithm include: (1) setting virtual detection lines to extract vehicle feature points in the road section, (2) grouping these points into different vehicles by using the Adaboost classifier, and (3) tracking the points and extracting vehicle trajectories by means of the improved optical flow method. After that, the velocities and accelerations of vehicles are calculated from the obtained trajectory data. Finally, vehicle emission in the specified road section is estimated with the help of the relationship between VSP and base emission rate. Experiments were conducted in a highway section, and the result validated the feasibility of the method.
Vehicle Emission Estimation Based on Video Tracking of Vehicle Trajectories
17th COTA International Conference of Transportation Professionals ; 2017 ; Shanghai, China
CICTP 2017 ; 2942-2951
2018-01-18
Conference paper
Electronic Resource
English
Vehicle trajectories evaluation by static video sensors
IEEE | 2006
|Reconstructing Vehicle Trajectories to Support Travel Time Estimation
Transportation Research Record | 2018
|Estimation of Vehicle Trajectories with Locally Weighted Regression
Online Contents | 2007
|Estimation of Vehicle Trajectories with Locally Weighted Regression
Transportation Research Record | 2007
|