Video-Capture-Based Approach to Extract Multiple Vehicular Trajectory Data for Traffic Modeling

Abstract

This paper presents a methodology for extracting vehicular trajectory data using a video-capture technique, and describes a computer-based tool, Vehicle Video-Capture Data Collector (VEVID), which was developed to help extract trajectory data from the video. The proposed methodology consists of three basic steps. First, an urban street is videotaped from an elevated position, and distances between reference points are measured and input into the VEVID parameters file. Second, a segment of video is digitized into a Video for Windows (AVI) file at a user-specified frame rate. Third, the AVI file is registered in VEVID, and then, in each frame, the user simply clicks the mouse over a distinguishable point of targeted vehicles. Trajectories are output into a trajectory data file along with speeds, accelerations, and gaps of targeted vehicles. Finally, several types of data extracted from VEVID that are applicable to microscopic traffic simulation modeling on urban street networks are discussed to show the capabilities of the proposed approach in traffic modeling.