University of Twente Student Theses


Improving accuracy of vehicle tracking by fusing IMU and GPS

Ravichandran, R. (2023) Improving accuracy of vehicle tracking by fusing IMU and GPS.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Embargo date:21 February 2028
Abstract:The accurate tracking of vehicles is a crucial aspect in the field of telematics. GPS, which is widely used for vehicle tracking, can be influenced by various factors such as signal interference, atmospheric conditions, and receiver noise, making it difficult to achieve accurate tracking results. In addition, GPS signals can become weak or even unavailable in certain environments, leading to further challenges in tracking accuracy. To overcome these limitations, a combination of GPS data and Inertial Measurement Unit (IMU) sensor data can be used to improve the accuracy of vehicle tracking. IMU sensors have a higher sample rate compared to GPS and provide more precise position data, but they are also prone to noise. Therefore, calibration is done to remove sensor noise. The Extended Kalman Filter (EKF) is used to fuse data from various sources, including the accelerometer, gyroscope, and magnetometer, along with GPS data to perform the dead reckoning process. The EKF combines multiple sources of information to produce robust and reliable tracking data, even in challenging environments. Madgwick’s filter is computationally more efficient than the Kalman filter to estimate the orientation, which is then used as input to the Extended Kalman Filter. Over time, errors can accumulate in the dead reckoning process, and calibration of the dead reckoning algorithm using reference points such as speed bumps or road landmarks are performed to improve accuracy. In this thesis, speed bumps are used as a reference point. The proposed method demonstrates improved accuracy in vehicle tracking even when GPS data is not available for a few seconds and when the vehicle is stopped. Additionally, the jumping of GPS positions can be eliminated using the proposed method.
Item Type:Essay (Master)
Capte, Amsterdam, Netherlands
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Embedded Systems MSc (60331)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page