Real-time vehicle detection for traffic monitoring by applying a deep learning algorithm over images acquired from satellite and drone
International Journal of Intelligent Unmanned Systems
ISSN: 2049-6427
Article publication date: 19 September 2022
Issue publication date: 30 October 2023
Abstract
Purpose
The purpose is to design a system in which drones can control traffic most effectively using a deep learning algorithm.
Design/methodology/approach
Drones have now started entry into each facet of life. The entry of drones has made them a subject of great relevance in the present technological era. The span of drones is, however, very broad due to various kinds of usages leading to different types of drones. Out of the many usages, one usage which is presently being widely researched is traffic monitoring as traffic monitoring can hover over a particular area. This paper specifically brings out the basic algorithm You Look Only Once (YOLO) which may be used for identifying the vehicles. Consequently, using deep learning YOLO algorithm, identification of vehicles will, therefore, help in easy regulation of traffic in streetlights, avoiding accidents, finding out the culprit drivers due to which traffic jam would have taken place and recognition of a pattern of traffic at various timings of the day, thereby announcing the same through radio (namely, Frequency Modulation (FM)) channels, so that people can take the route which is the least jammed.
Findings
The study found that the object(s) detected by the deep learning algorithm is almost the same as if seen from a naked eye from the top view. This led to the conclusion that the drones may be used for traffic monitoring, in the days to come, which was not the case earlier.
Originality/value
The main research content and key algorithm have been introduced. The research is original. None of the parts of this research paper has been published anywhere.
Keywords
Citation
Vohra, D.S., Garg, P.K. and Ghosh, S. (2023), "Real-time vehicle detection for traffic monitoring by applying a deep learning algorithm over images acquired from satellite and drone", International Journal of Intelligent Unmanned Systems, Vol. 11 No. 4, pp. 441-452. https://doi.org/10.1108/IJIUS-06-2022-0077
Publisher
:Emerald Publishing Limited
Copyright © 2022, Emerald Publishing Limited