Statistical accuracy of an extraction algorithm for linear image objects

dc.contributor.advisorFabris-Rotelli, Inger Nicolette
dc.contributor.coadvisorStein, Alfred
dc.contributor.coadvisorDebba, Pravesh
dc.contributor.emailrenate.thiede@gmail.comen_ZA
dc.contributor.postgraduateThiede, Renate Nicole
dc.date.accessioned2020-02-11T13:02:13Z
dc.date.available2020-02-11T13:02:13Z
dc.date.created2020
dc.date.issued2019
dc.descriptionMini Dissertation (MSc)--University of Pretoria, 2019.en_ZA
dc.description.abstractInformal unpaved roads in developing countries arise naturally through human movement and informal housing setups. These roads are not authorised nor maintained by council, nor recorded in official databases or online maps. Mapping such roads from satellite images is a common problem, as information on these roads is critical for sustainable city growth. Information on their location and extent may be gleaned from spatial big data, however, no automatic or semi-automatic approach is freely available. This research develops a novel algorithm for extracting informal roads from multispectral satellite images, using physical road characteristics. These include near-infrared reflectance, addressed via the NDVI index, shape, addressed via measures of compactness and elongation, and grey-value intensity. The crux of the algorithm is the Discrete Pulse Transform, implemented via the Roadmaker's Pavage. The algorithm provides a classification of road objects, along with an associated uncertainty measure for each road object. Accuracy is assessed using per-pixel assessment metrics and metrics based on road characteristics, including completeness, correctness, and Pratt's Figure of Merit, which is applied to road extraction accuracy for the first time. The algorithm is applied to areas in Gauteng and North West Provinces, South Africa. Sources of uncertainty and error are discussed, such as indefinite boundaries, surface type heterogeneity, trees and shadows.en_ZA
dc.description.availabilityUnrestricteden_ZA
dc.description.degreeMScen_ZA
dc.description.departmentStatisticsen_ZA
dc.description.sponsorshipAcknowledgement of the National Research Foundation for the funding provided through the NRF-SASA Crisis in Academic Statistics grant.en_ZA
dc.identifier.citationThiede, RN 2019, Statistical accuracy of an extraction algorithm for linear image objects, MSc Mini Dissertation, University of Pretoria, Pretoria, viewed yymmdd <http://hdl.handle.net/2263/73211>en_ZA
dc.identifier.otherA2020en_ZA
dc.identifier.urihttp://hdl.handle.net/2263/73211
dc.language.isoenen_ZA
dc.publisherUniversity of Pretoria
dc.rights© 2019 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria.
dc.subjectUCTDen_ZA
dc.subjectMathematical Statisticsen_ZA
dc.titleStatistical accuracy of an extraction algorithm for linear image objectsen_ZA
dc.typeMini Dissertationen_ZA

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Thiede_Statistical_2019.pdf
Size:
39.94 MB
Format:
Adobe Portable Document Format
Description:
Mini Dissertation

License bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
1.75 KB
Format:
Item-specific license agreed upon to submission
Description: