Deep Learning Based Vehicle Speed Estimation on Highways

Traffic management always is a matter requiring the attention of highway system managers in terms vehicle monitoring and speed estimation. This paper proposes an efficient deep learning-based estimation on lanes Vietnam transport system. The input videos are recorded by fixed surveillance cameras. An optimized single shot multibox detector network, called SSD, utilized for license plate detection (LPD). SORT (simple online real-time tracking) model first applied to video tracking performed detected area. process investigates traveling distance estimate its speed. In this study, dataset has been normalized improve efficiency localization time elapsing traveled vehicles highways. results showed that proposed achieved better accuracy determined speeds with errors ranging between [-1.5, +1.1] km/h, equivalent 98% error limit regulation Viet Nam.