Correction of Very High-Resolution Satellite Images using Control Points Captured by Web Map Service (WMS) server: Google Earth
Ashraf M. A. Shrawai1, Ateaya B. Azeez2
1Ashraf M. A. Shrawai*, Researcher, Aerial Photography Section, Aerial Photography & Aviation Dep, National Authority for Remote Sensing & Space Sciences, Cairo, Egypt.
2Ateaya B. Azeez, Researcher, Ground Surveying Section, Aerial Photography & Aviation Dep., National Authority for Remote Sensing & Space Sciences, Cairo, Egypt.
Manuscript received on May 01, 2020. | Revised Manuscript received on May 17, 2020. | Manuscript published on June 10, 2020. | PP: 188-194 | Volume-9 Issue-8, June 2020. | Retrieval Number: 100.1/ijitee.H6267069820 | DOI: 10.35940/ijitee.H6267.069820
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The recent progress for spatial resolution of remote sensing imagery led to generate many types of Very High-Resolution (VHR) satellite images, consequently, general speaking, it is possible to prepare accurate base map larger than 1:10,000 scale. One of these VHR satellite image is WorldView-3 sensor that launched in August 2014. The resolution of 0.31m makes WorldView-3 the highest resolution commercial satellite in the world. In the current research, a pan-sharpen image from that type, covering an area at Giza Governorate in Egypt, used to determine the suitable large-scale map that could be produced from that image. To reach this objective, two different sources for acquiring Ground Control Points (GCPs). Firstly, very accurate field measurements using GPS and secondly, Web Map Service (WMS) server (in the current research is Google Earth) which is considered a good alternative when GCPs are not available, are used. Accordingly, three scenarios are tested, using the same set of both 16 Ground Control Points (GCPs) as well as 14 Check Points (CHKs), used for evaluation the accuracy of geometric correction of that type of images. First approach using both GCPs and CHKs coordinates acquired by GPS. Second approach using GCPs coordinates acquired by Google Earth and CHKs acquired by GPS. Third approach using GCPs and CHKs coordinates by Google Earth. Results showed that, first approach gives Root Mean Square Error (RMSE) planimeteric discrepancy for GCPs of 0.45m and RMSE planimeteric discrepancy for CHKs of 0.69m. Second approach gives RMSE for GCPs of 1.10m and RMSE for CHKs of 1.75m. Third approach gives RMSE for GCPs of 1.10m and RMSE for CHKs of 1.40m. Taking map accuracy specification of 0.5mm of map scale, the worst values for CHKs points (1.75m&1,4m) resulted from using Google Earth as a source, gives the possibility of producing 1:5000 large-scale map compared with the best value of (0.69m) (map scale 1:2500). This means, for the given parameters of the current research, large scale maps could be produced using Google Earth, in case of GCPs are not available accurately from the field surveying, which is very useful for many users.
Keywords: WorldView-3, Very high-resolution satellite images, Geometric correction, GPS, Web Map Service (WMS) server, Google Earth.
Scope of the Article: Web Mining