For a scene with “non-groundlocked/ ground-control:true” the band alignment will be low. Scene-to-scene alignment may provide good band alignment, but in many cases it will not.
A ground control point (GCP) is feature that you can clearly identify in the raw image for which you have a known ground coordinate.They are used in the creation of every image and the accuracy of the product will vary from region to region based on available GCPs.
The locational accuracy of all the imagery products depends on the quality of the reference data used: Ground Control Points (GCPs) and Digital Elevation Model (DEMs). Additionally, the roll angle of the spacecraft during the image acquisition and the number as well as the distribution of GCPs within the image will impact the final product accuracy.
To ensure the high accuracy of all of our ortho products on a global basis, Planet uses GCPs derived from high resolution satellite and airborne imagery. For most of Earth’s land mass, GCPs are derived from an ALOS 2.5 m resolution basemap. NAIP is used over the continental United States and Landsat 8 is used as a fallback solution over remote polar areas and some small islands.
The vertical component is derived from DEMs with a post spacing under 30 m globally. Planet products produced using GCPs and the World30 DEM will have a locational accuracy of 10 m RMSE or better. Internal testing conducted on multiple locations worldwide indicates that locational accuracy will typically (80% of the times) be better than 7 m RMSE.
The percentage of SkySat frames that make up the full Collect product that have good ground control.
One of the metadata fields of an image is called “ground_control”, which is a boolean (true or false value). True, corresponds to the image passing all GCP checks, whereas as false corresponds to image not able to get enough GCPs, leading to the image not being geometrically aligned.