Document Type Doctoral Thesis Author Salmon, Brian Paxton URN etd-09252012-174827 Document Title Improved hyper-temporal feature extraction methods for land cover change detection in satellite time series Degree PhD(Eng) Department Electrical, Electronic and Computer Engineering Supervisor
Advisor Name Title Dr F van den Bergh Co-Supervisor Prof J C Olivier Supervisor Keywords
- time series
- Fourier transform
- change detection
- extended Kalman filter
Date 2012-09-06 Availability unrestricted AbstractThe growth in global population inevitably increases the consumption of natural resources. The need to provide basic services to these growing communities leads to an increase in anthropogenic changes to the natural environment. The resulting transformation of vegetation cover (e.g. deforestation, agricultural expansion, urbanisation) has significant impacts on hydrology, biodiversity, ecosystems and climate. Human settlement expansion is the most common driver of land cover change in South Africa, and is currently mapped on an irregular, ad hoc basis using visual interpretation of aerial photographs or satellite images. This thesis proposes several methods of detecting newly formed human settlements using hyper-temporal, multi-spectral, medium spatial resolution MODIS land surface reflectance satellite imagery. The hyper-temporal images are used to extract time series, which are analysed in an automated fashion using machine learning methods. A post-classification change detection framework was developed to analyse the time series using several feature extraction methods and classifiers. Two novel hyper-temporal feature extraction methods are proposed to characterise the seasonal pattern in the time series. The first feature extraction method extracts Seasonal Fourier features that exploits the difference in temporal spectra inherent to land cover classes. The second feature extraction method extracts state-space vectors derived using an extended Kalman filter. The extended Kalman filter is optimised using a novel criterion which exploits the information inherent in the spatio-temporal domain. The post-classification change detection framework was evaluated on different classifiers; both supervised and unsupervised methods were explored. A change detection accuracy of above 85% with false alarm rate below 10% was attained. The best performing methods were then applied at a provincial scale in the Gauteng and Limpopo provinces to produce regional change maps, indicating settlement expansion.
© 2012 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.
Please cite as follows:
Salmon, BP 2012, Improved hyper-temporal feature extraction methods for land cover change detection in satellite time series, PhD thesis, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-09252012-174827/ >
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access 00front.pdf 70.92 Kb 00:00:19 00:00:10 00:00:08 00:00:04 < 00:00:01 01chapters1-2.pdf 2.25 Mb 00:10:26 00:05:22 00:04:41 00:02:20 00:00:12 02chapter3.pdf 1.24 Mb 00:05:44 00:02:57 00:02:35 00:01:17 00:00:06 03chapter4.pdf 904.71 Kb 00:04:11 00:02:09 00:01:53 00:00:56 00:00:04 04chapter5.pdf 948.39 Kb 00:04:23 00:02:15 00:01:58 00:00:59 00:00:05 05chapter6.pdf 250.86 Kb 00:01:09 00:00:35 00:00:31 00:00:15 00:00:01 06chapter7.pdf 171.47 Kb 00:00:47 00:00:24 00:00:21 00:00:10 < 00:00:01 07chapter8.pdf 1.58 Mb 00:07:18 00:03:45 00:03:17 00:01:38 00:00:08 08chapter9.pdf 107.73 Kb 00:00:29 00:00:15 00:00:13 00:00:06 < 00:00:01 09back.pdf 185.26 Kb 00:00:51 00:00:26 00:00:23 00:00:11 < 00:00:01