Vehicle trajectory prediction is a crucial task in the field of intelligent transportation systems, with applications in traffic management, autonomous driving, and advanced driver assistance systems. In this project, we aim to demonstrate the capabilities of this technology by collecting and analyzing video footage of traffic in Thu Duc District, Ho Chi Minh city. Utilizing YOLOv7 for object detection and DeepSORT for object tracking, we are able to accurately gather the coordinates of vehicles in the scene. Using these coordinates, we then employ various hybrid models such as CNN-LSTM and CNN-GRU to predict the future trajectory of the vehicles. The results are then visualized and superimposed onto the original video footage to showcase the capabilities of these predictive models.
Leverage Deep Learning Methods for Vehicle Trajectory Prediction in Chaotic Traffic
Lect. Notes on Data Eng. and Comms.Technol.
International Conference on Intelligence of Things ; 2023 ; Ho Chi Minh City, Vietnam October 25, 2023 - October 27, 2023
20.10.2023
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Network-Wide Vehicle Trajectory Prediction in Urban Traffic Networks using Deep Learning
Transportation Research Record | 2018
|Elsevier | 2024
|Trajectory Data-Driven Network Representation for Traffic State Prediction using Deep Learning
Springer Verlag | 2024
|Traffic density adaptive vehicle trajectory prediction method
Europäisches Patentamt | 2024
|