In an area of interest- Sivas Basin, Turkey- where most of the units are sedimentary and show similar spectral characteristics, spectral settings of ASTER sensor may not be enough by itself. Therefore, considering other aspects, such as morphological variables, is reasonable in addition to spectral classifiers. The main objective of this study is to test usefulness of integration of spectral analysis and morphological information for geological mapping. Remotely sensed imagery obtained from ASTER sensor is used to classify different lithological units while DEM is used to characterize landforms related to these lithological units.
Maximum Likelihood Classification (MLC) is used to integrate data streaming from different sources. The methodology involves integrating the surface properties of the classified geological units in addition to the spectral reflectances. Seven different classification trials were conducted: : 1. MLC using only nine ASTER bands, 2. MLC using ASTER bands and DEM, 3. MLC using ASTER bands and slope, 4. MLC using ASTER bands and plan curvature, 5. MLC using ASTER bands and profile curvature, 6. MLC using ASTER bands and drainage density and finally 7. MLC using ASTER bands and all ancillary data.
The results revealed that integrating topographical parameters aid in improvement of classification where spectral information is not sufficient to discriminate between classes of interest. An increase of more than 5% is observed in overall accuracy for the all ancillary data integration case. Moreover more than 10% improvement for most of the classes was identified. However from the results it is evident that the areal extent of the classified units causes constraints on application of the methodology.
| Identifer | oai:union.ndltd.org:METU/oai:etd.lib.metu.edu.tr:http://etd.lib.metu.edu.tr/upload/3/12610626/index.pdf |
| Date | 01 May 2009 |
| Creators | Akkok, Inci |
| Contributors | Suzen, Lutfi Mehmet |
| Publisher | METU |
| Source Sets | Middle East Technical Univ. |
| Language | English |
| Detected Language | English |
| Type | M.S. Thesis |
| Format | text/pdf |
| Rights | To liberate the content for public access |
Page generated in 0.0019 seconds