Accuracy estimates of digital elevation model (DEM) towards the creation of satellite image map (SIM) / Kamaruzzaman Wahid

Satellite Image Map (SIM) integrates Remote Sensing, Geographical Information System (GIS), Global Positioning System (GPS) and Cartography in the production of digital cartographic maps. This thesis, explores the methodology towards evaluating the accuracy and effects of the use of digital elevation model (DEM) towards the creation of SIM. The DEM which were constructed using the topographical contour data, LiDAR and SPOT 5 HRG (High Resolution Geometric) were evaluated quantitatively and qualitatively. Quantitative evaluation were carried out using the Root Mean Square Error (RMSE) which are calculated based on the discrepancies of absolute and horizontal values of generated results toward the well defined point determined in the study. In the qualitative evaluation, the mismatch between the topographical, LiDAR and SPOT 5 HRG data were evaluated using four (4) different methods namely; quality evaluation of the DEM generated, orthorectification results, scale comparisons and comparison between the SPOT 5 HRG data with the SIM product produced by MACRES using the Landsat TM data. The quantitative results shows that the RMSEz for Topographical, LiDAR and SPOT 5 HRG are 1.408 m, 0.250 m and 6.872 m respectively. The RMSEz for Topographical LiDAR complies with the USGS Mapping Accuracy Standards (1947) for mapping scale of 1:10 000 and above. However, SPOT 5 HRG complies with the USGS Mapping Accuracy Standards (1947) of 1: 25,000 scale and above respectively. For horizontal accuracy the RMSExy for the Topographical, LiDAR and SPOT 5 HRG were 1.295 m, 0.941 m and 2.481 m respectively. For the qualitative results, topographical data provides a smooth generated DEM. The combination of the DEM with the 2.5 m SPOT 5 HRG data presents a good orthorectification results for the creation of SIM.