Document Type
Technical Report
Publication Date
12-1-2023
Keywords
UAV, unmanned aircraft system, drones, traffic, transportation, incident management, UAS applications, object detection algorithm, experimental design
Digital Object Identifier (DOI)
https://doi.org/10.5038/CUTR-NICR-Y2-4-4.2
Abstract
In the second phase of the project, the research team continued traffic data collection with unmanned aerial systems (UAS) and dual sensors (RGB and thermal cameras) for roadways under normal operational condition and during incident clearance. The research team conducted a thorough literature review of incident detection methods. Based on the review outcomes, the research team extracted traffic features from learning the video data and trained several learning models for identifying non-congestion conditions caused by incidents. This method was tested on a two-minute video with data captured by a drone at a location at which traffic was passing through an incident site. The results show that some machine learning models (support vector machine, K nearest neighbor, and random forest) performed very well in F1 scoring.
Scholar Commons Citation
Zhang, Yu; Kourtellis, Achilleas; Post, Joseph; Tang, Hualong; Yelchuri, Keerthana; and Porter, Brian, "Corridor-Wide Surveillance Using Unmanned Aircraft Systems" (2023). Research Reports. 24.
https://digitalcommons.usf.edu/cutr_nicr/24
Policy Brief
