dc.contributor.author |
Wessels, Konrad J
|
|
dc.contributor.author |
Van den Bergh, F
|
|
dc.contributor.author |
Roy, DP
|
|
dc.contributor.author |
Salmon, BP
|
|
dc.contributor.author |
Steenkamp, Karen C
|
|
dc.contributor.author |
MacAlister, B
|
|
dc.contributor.author |
Swanepoel, D
|
|
dc.contributor.author |
Jewit, D
|
|
dc.date.accessioned |
2017-02-23T10:02:12Z |
|
dc.date.available |
2017-02-23T10:02:12Z |
|
dc.date.issued |
2016 |
|
dc.identifier.citation |
Wessels, K.J., Van den Bergh, F., Roy, D.P., Salmon, B.P., Steenkamp, K.C., MacAlister, B., Swanepoel, D. and Jewit, D. 2016. Rapid land cover map updates using change detection and robust random forest classifiers. Remote Sensing, 8(11), 1-24 |
en_US |
dc.identifier.issn |
2072-4292 |
|
dc.identifier.uri |
http://www.mdpi.com/2072-4292/8/11/888
|
|
dc.identifier.uri |
http://hdl.handle.net/10204/8964
|
|
dc.description |
Copyright: 2016 MDPI |
en_US |
dc.description.abstract |
The paper evaluated the Landsat Automated Land Cover Update Mapping (LALCUM) system designed to rapidly update a land cover map to a desired nominal year using a pre-existing reference land cover map. The system uses the Iteratively Reweighted Multivariate Alteration Detection (IRMAD) to identify areas of change and no change. The system then automatically generates large amounts of training samples (n > 1 million) in the no-change areas as input to an optimized Random Forest classifier. Experiments were conducted in the KwaZulu-Natal Province of South Africa using a reference land cover map from 2008, a change mask between 2008 and 2011 and Landsat ETM+ data for 2011. The entire system took 9.5 h to process. We expected that the use of the change mask would improve classification accuracy by reducing the number of mislabeled training data caused by land cover change between 2008 and 2011. However, this was not the case due to exceptional robustness of Random Forest classifier to mislabeled training samples. The system achieved an overall accuracy of 65%–67% using 22 detailed classes and 72%–74% using 12 aggregated national classes. “Water”, “Plantations”, “Plantations—clearfelled”, “Orchards—trees”, “Sugarcane”, “Built-up/dense settlement”, “Cultivation—Irrigated” and “Forest (indigenous)” had user’s accuracies above 70%. Other detailed classes (e.g., “Low density settlements”, “Mines and Quarries”, and “Cultivation, subsistence, drylands”)which are required for operational, provincial-scale land use planning and are usually mapped using manual image interpretation, could not be mapped using Landsat spectral data alone. However, the system was able to map the 12 national classes, at a sufficiently high level of accuracy for national scale land cover monitoring. This update approach and the highly automated, scalable LALCUM system can improve the efficiency and update rate of regional land cover mapping. |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
MDPI |
en_US |
dc.relation.ispartofseries |
Wokflow;17918 |
|
dc.subject |
Landsat |
en_US |
dc.subject |
Land cover |
en_US |
dc.subject |
Change detection |
en_US |
dc.subject |
Automated mapping |
en_US |
dc.subject |
Random forest |
en_US |
dc.title |
Rapid land cover map updates using change detection and robust random forest classifiers |
en_US |
dc.type |
Article |
en_US |
dc.identifier.apacitation |
Wessels, K. J., Van den Bergh, F., Roy, D., Salmon, B., Steenkamp, K. C., MacAlister, B., ... Jewit, D. (2016). Rapid land cover map updates using change detection and robust random forest classifiers. http://hdl.handle.net/10204/8964 |
en_ZA |
dc.identifier.chicagocitation |
Wessels, Konrad J, F Van den Bergh, DP Roy, BP Salmon, Karen C Steenkamp, B MacAlister, D Swanepoel, and D Jewit "Rapid land cover map updates using change detection and robust random forest classifiers." (2016) http://hdl.handle.net/10204/8964 |
en_ZA |
dc.identifier.vancouvercitation |
Wessels KJ, Van den Bergh F, Roy D, Salmon B, Steenkamp KC, MacAlister B, et al. Rapid land cover map updates using change detection and robust random forest classifiers. 2016; http://hdl.handle.net/10204/8964. |
en_ZA |
dc.identifier.ris |
TY - Article
AU - Wessels, Konrad J
AU - Van den Bergh, F
AU - Roy, DP
AU - Salmon, BP
AU - Steenkamp, Karen C
AU - MacAlister, B
AU - Swanepoel, D
AU - Jewit, D
AB - The paper evaluated the Landsat Automated Land Cover Update Mapping (LALCUM) system designed to rapidly update a land cover map to a desired nominal year using a pre-existing reference land cover map. The system uses the Iteratively Reweighted Multivariate Alteration Detection (IRMAD) to identify areas of change and no change. The system then automatically generates large amounts of training samples (n > 1 million) in the no-change areas as input to an optimized Random Forest classifier. Experiments were conducted in the KwaZulu-Natal Province of South Africa using a reference land cover map from 2008, a change mask between 2008 and 2011 and Landsat ETM+ data for 2011. The entire system took 9.5 h to process. We expected that the use of the change mask would improve classification accuracy by reducing the number of mislabeled training data caused by land cover change between 2008 and 2011. However, this was not the case due to exceptional robustness of Random Forest classifier to mislabeled training samples. The system achieved an overall accuracy of 65%–67% using 22 detailed classes and 72%–74% using 12 aggregated national classes. “Water”, “Plantations”, “Plantations—clearfelled”, “Orchards—trees”, “Sugarcane”, “Built-up/dense settlement”, “Cultivation—Irrigated” and “Forest (indigenous)” had user’s accuracies above 70%. Other detailed classes (e.g., “Low density settlements”, “Mines and Quarries”, and “Cultivation, subsistence, drylands”)which are required for operational, provincial-scale land use planning and are usually mapped using manual image interpretation, could not be mapped using Landsat spectral data alone. However, the system was able to map the 12 national classes, at a sufficiently high level of accuracy for national scale land cover monitoring. This update approach and the highly automated, scalable LALCUM system can improve the efficiency and update rate of regional land cover mapping.
DA - 2016
DB - ResearchSpace
DP - CSIR
KW - Landsat
KW - Land cover
KW - Change detection
KW - Automated mapping
KW - Random forest
LK - https://researchspace.csir.co.za
PY - 2016
SM - 2072-4292
T1 - Rapid land cover map updates using change detection and robust random forest classifiers
TI - Rapid land cover map updates using change detection and robust random forest classifiers
UR - http://hdl.handle.net/10204/8964
ER -
|
en_ZA |