Artificial Intelligence and Communication Technologies
Vehicle Speed Estimation using Object Detection for Intelligent Traffic Management
Authors: Atharva Hiwarekar, Swaroop Chavhan, Onkar Deshpande and Vedant Joshi
Publishing Date: 18-02-2023
ISBN: 978-81-955020-5-9
Abstract
Effective and safe transport is an essential need of every individual, and it plays a vital role in every aspect of our life. Traffic Management has become a key challenge in today’s developing world. Managing growing surveillance in today’s world is a challenge to the administration, which can be solved by innovations in technology. Over-speeding is one of the main factors that cause road accidents, followed by injuries and deaths. Thus there is a need to develop a solution that will detect traffic and calculate vehicle speed in real-time. This paper deals with solving traffic problems using Object Detection. We used YOLOv3 and Deep-SORT algorithm to detect and track vehicles from input surveillance video. And, we calculate the speed of the vehicles present in the input surveillance video using the Frame difference method. We have shown that object detection can be used to identify and track vehicles in extreme weather conditions like rain and snow. Further, it is shown that the frame difference method and linear perspective transformation can be used to calculate vehicle speed with RMSE as low as 4.568 km/hr.
Keywords
Machine learning, Deep Learning, Computer Vision, Object Detection
Cite as
Atharva Hiwarekar, Swaroop Chavhan, Onkar Deshpande and Vedant Joshi, "Vehicle Speed Estimation using Object Detection for Intelligent Traffic Management", In: Saroj Hiranwal and Garima Mathur (eds), Artificial Intelligence and Communication Technologies, SCRS, India, 2023, pp. 677-685. https://doi.org/10.52458/978-81-955020-5-9-64