Enhanced Vision Based Multi-Sensor Positioning Using Nonlinear Error Modeling

dc.contributor.authorRagab, Mahmoud
dc.contributor.departmentElectrical and Computer Engineering
dc.contributor.supervisorNoureldin, Aboelmagd
dc.contributor.supervisorGivigi, Sidney
dc.degree.grantorQueen's University at Kingstonen
dc.description.abstractPositioning and navigation (POS/NAV) systems are essential for future land vehicles. These systems have become standard in many of the new automobile models. The design of POS/NAV systems has evolved to include functionalities that make them more than just a vehicle location and map displaying system. Future land vehicle POS/NAV technology will integrate several sensors and systems to increase driving efficiency and enhance driver experience and safety. The core of these systems that provide drivers with route guidance information is the Global Navigation Satellite System (GNSS). GNSS positioning accuracy in dense urban canyon environments significantly deteriorates due to multipath and signal blockage. For this reason, the inertial navigation system (INS) is usually integrated with GNSS to guarantee a trustworthy navigation solution during such periods of GNSS signal outages. A low-cost navigation solution for land vehicles has been established by integrating the GNSS positioning solution with the measurements from the vehicle strapdown inertial sensors (accelerometers and gyroscopes). The main shortcoming of utilizing inertial sensors is the gradual error accumulation, where the gyroscope drift errors increase progressively, leading to an impractical position estimate, particularly without GNSS updates. Thus, navigation in any GNSS-denied environment necessitates aiding INS with other exteroceptive sensors such as cameras through visual odometry (VO) to ensure efficient and reliable positioning updates. The aim of this research is to develop an integrated multi-sensor POS/NAV system for land vehicles capable of offering seamless positioning at metre level accuracy. The core development will be founded on a VO-based multi-sensor fusion system to surmount the inaccuracy of GNSS in urban areas and motion sensors' drift. Moreover, to improve the overall system accuracy of the VO-based navigation solution, we propose an integration scheme that enhances the positioning accuracy during GNSS outages by nonlinear modeling of the residual position errors using the fast orthogonal search (FOS) algorithm. The proposed system is assessed on several real road trajectories with different motion dynamics, driving situations, and prolonged GNSS outages. GNSS outages of up to 10 minutes were intentionally simulated to examine the overall system performance. The results show a significant improvement in the positioning accuracy that can reach around 80% when compared to the current methods that rely on only integrating VO with inertial sensor and GNSS.
dc.embargo.termsI want to restrict my thesis for 2-years as I am still in the process of publishing journal papers that are related to my thesis.
dc.relation.ispartofseriesCanadian thesesen
dc.rightsAttribution 4.0 International*
dc.rightsQueen's University's Thesis/Dissertation Non-Exclusive License for Deposit to QSpace and Library and Archives Canadaen
dc.rightsProQuest PhD and Master's Theses International Dissemination Agreementen
dc.rightsIntellectual Property Guidelines at Queen's Universityen
dc.rightsCopying and Preserving Your Thesisen
dc.rightsThis publication is made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws without written authority from the copyright owner.en
dc.subjectGlobal Navigation Satellite Systems
dc.subjectKalman Filter
dc.subjectSensor Fusion
dc.subjectVisual Odometry
dc.subjectNonlinear Error Modeling
dc.subjectInertial Navigation
dc.titleEnhanced Vision Based Multi-Sensor Positioning Using Nonlinear Error Modeling
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Ragab_Mahmoud_M_202308_PhD_Mahmoud Ragab.pdf
9.43 MB
Adobe Portable Document Format