Remote Sensing #10 - Image Corrections

The theory behind performing image corrections to correctly Georeference images. To do this we must line up the Datum, Projection and atmospheric and reflectivity of the day of the object being sensed. Be in mind of sun glint, shadowing, water depth etc... The Ground control points must be no greater than '1' for the ERROR (Root mean square error because "recall that RMSE has the same unit as the dependent variable (DV). It means that there is no absolute good or bad threshold, however you can define it based on your DV. For a datum which ranges from 0 to 1000, an RMSE of 0.7 is small, but if the range goes from 0 to 1, it is not that small anymore. However, although the smaller the RMSE, the better, you can make theoretical claims on levels of the RMSE by knowing what is expected from your DV in your field of research." Radiometric correction easy to understand website: http://gsp.humboldt.edu/OLM/Courses/GSP_216_Online/lesson4-1/radiometric.html This correction accounts for scattering of light (additive path radiance). Essentially an image once first captured is some random value of between 0-255, radiometric correction will split this number into meaningful units. However these units will still account for atmospheric rebounds of light such as by clouds (this is the additive path). Atmospheric correction will then remove this. Look more in-depth at this using ArcGIS Desktop website found here: https://pro.arcgis.com/en/pro-app/help/data/imagery/overview-of-georeferencing.htm Get the link to the textbook/PDF (Free!) for this course, get it here: https://www.crcsi.com.au/history-2/earth-observation-series-2/ Download volumes Vol 1A & and Vol 1B