Article: Real-time position and trajectory estimation based on deep learning and monocular cameras Journal: International Journal of Vehicle Safety (IJVS) 2023 Vol.13 No.1 pp.64 - 89 Abstract: Estimating the real-time state of a vehicle is critical for driver assistance systems. In this study, we propose a framework that utilises deep learning (DL) and monocular cameras to obtain information about vehicles. We first estimate the velocity of the ego vehicle using optical flow. Then, the position of the visible vehicle is tracked and predicted using deepSORT and projection geometry. Finally, an extended Kalman filter (EKF) is used to combine the velocity, tracking information and relative position to estimate the trajectory of both vehicles in a common coordinate frame. We evaluated the accuracy of the visible vehicle's position using three public data sets. The results show that our proposed algorithm has a consistent Root Mean Square Error (RMSE) between 5.12 and 6.272. On two of the data sets, our algorithm outperforms other data-driven supervised DDSDL algorithms. To evaluate trajectory generation, we conducted both simulated experiments and real-world tests, demonstrating that our algorithm generates accurate vehicle trajectories. Inderscience Publishers - linking academia, business and industry through research

Title: Real-time position and trajectory estimation based on deep learning and monocular cameras

Authors: Lianen Qu; Zhongwei Qu; Zhikao Ren

Addresses: College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao, Shandong, China; Sino-Thai International Rubber College, Qingdao University of Science and Technology, Qingdao, Shandong, China ' Sino-Thai International Rubber College, Qingdao University of Science and Technology, Qingdao, Shandong, China ' College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao, Shandong, China

Abstract: Estimating the real-time state of a vehicle is critical for driver assistance systems. In this study, we propose a framework that utilises deep learning (DL) and monocular cameras to obtain information about vehicles. We first estimate the velocity of the ego vehicle using optical flow. Then, the position of the visible vehicle is tracked and predicted using deepSORT and projection geometry. Finally, an extended Kalman filter (EKF) is used to combine the velocity, tracking information and relative position to estimate the trajectory of both vehicles in a common coordinate frame. We evaluated the accuracy of the visible vehicle's position using three public data sets. The results show that our proposed algorithm has a consistent Root Mean Square Error (RMSE) between 5.12 and 6.272. On two of the data sets, our algorithm outperforms other data-driven supervised DDSDL algorithms. To evaluate trajectory generation, we conducted both simulated experiments and real-world tests, demonstrating that our algorithm generates accurate vehicle trajectories.

Keywords: deep learning; relative position estimation; trajectory generation; non-linear optimisation.

DOI: 10.1504/IJVS.2023.137693

International Journal of Vehicle Safety, 2023 Vol.13 No.1, pp.64 - 89

Received: 05 May 2023
Accepted: 30 Nov 2023
Published online: 02 Apr 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Real-time position and trajectory estimation based on deep learning and monocular cameras

Keep up-to-date