Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.755820
Title: Simultaneous localisation and mapping for surveying applications
Author: Williams, Emily
Awarding Body: University of Nottingham
Current Institution: University of Nottingham
Date of Award: 2014
Availability of Full Text:
Full text unavailable from EThOS.
Please contact the current institution’s library for further details.
Abstract:
Mobile Laser Scanning (MLS) is a data capture technique that has transformed the way in which survey data is collected especially for infrastructure corridors such as highways and railways. Using an integrated GPS and Inertial Measuring Unit (IMU), the mobile vehicle and therefore the collected data has a proven accuracy of +/- 10mm, when using ground control points. This allows high accuracy data collection at normal highway speeds. The main limitation of this type of data capture is the reliance on GPS for positioning. Although blackouts in GPS signal can be bridged using the IMU, the resulting error follows a time dependent model resulting in large errors after short periods of time. This prohibits the use of MLS for high accuracy applications in areas such as indoors and underground. Simultaneous Localisation and Mapping (SLAM) is a robotics technique that collects data about the environment as it navigates in order to make deeisionson its movement creating a completely autonomous robot. The main advantage of SLAM for mapping is that the robot does not rely on GPS to navigate its surroundings. Instead it relies wholly on relative measurements to the environment in which it travels. The maps created are a biproduct of the navigation solution and the accuracy of them is often overlooked. In order to use the maps created for purposes other than autonomy it is important that their creation is investigated and the final maps are critically assessed. This thesis is an investigation into using SLAM for the creation of high accuracy maps. Analysis is undertaken to assess the ability of these maps to be used for surveying applications. Initially an investigation is undertaken to understand the processes involved for the production of SLAM maps. With the primary outcome being to determine the perceived benefits of using the maps created using SLAM techniques for mobile mapping applications. Trials are undertaken to critically evaluate the capability of typical sensors used on board SLAM systems for the production of reliable measurements. The Zebedee system, developed by CSIRO is identified as a mapping tool which uses SLAM techniques. An error model is developed for this system using a traditional survey approach, the results of which are compared against data collected in an indoor and an underground environment. The resulting point clouds from the test sites are assessed to determine the surveying applications which this type of system can provide data for.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.755820  DOI: Not available
Share: