Hi,
I am trying to carry out orthorectification of a pair of CORONA satellite images.
My processing steps are as below:
1. Create a new chunk in Agisoft and load the two overlapping images
2. Align photos
3. Create GCPs and add information on elevations
4. Optimize camera
5. Re-run align photos
6. Build dense cloud
7. Build DEM
8. Build orthophoto
I am using 40 well distributed Ground Control Points (GCPs) based on the coordinates taken from satellite images in the ArcGIS basemap and elevations from the 30m SRTM.
However, my orthoimage is not coming out very well - with misalignment is around 700 metres in some parts of the image although my total overall error of check points is relatively low (124 metres) and DEM and dense cloud looks really good.
I am not sure if I am missing out on any critical steps in my workflow but if anyone has any tips or advice on what I could do to improve the accuracy of my orthorectified image output.