Processing is essential for transforming raw geospatial data into meaningful imagery. Processing levels act as a framework for understanding the level of transformation and manipulation the data has undergone. To select the right data for your use case, it is important to understand the processing levels offered.
Optical imagery falls into three geometric processing levels: primary, georectified, and orthorectified.
Primary imagery is closest to the natural image acquired by the sensor. The resulting imagery isn’t georeferenced nor orthorectified, and is used as a key component of digital elevation model (DEM) generation and orthorectification. It accounts for distortions caused by the optics and electronics of the sensor, as well as its ephemeris.
Primary imagery allows for meaningful analysis while remaining close to the raw form of the data.
Georectified imagery is first calibrated and then georectified. Georectification adjusts imagery to align with an existing coordinate reference system. This ensures that features in the image correspond accurately to their geographic locations. There are two levels of georectification:
- Coarse georeferencing, that applies an average terrain height for the area of interest is applied across the image.
- Precise georeferencing, that applies terrain correction by referencing the image against a series of ground control points.
This is a mandatory step before using imagery in GIS software for accurate spatial analysis.
Orthorectified imagery is first georectified and then orthorectified. Orthorectification removes distortions caused by terrain variations, camera angle, and other factors. The resulting orthorectified image will appear as if it was captured from directly above the region.
This correction enables accurate measurements and comparisons in geographic analysis.
| Collection | Primary | Georectified | Orthorectified |
|---|---|---|---|
| AxelGlobe |  | |  Level 1C Level 2A |
| Beijing-3A | Level 1 with RPC | Level 2A | |
| Beijing-3N | Level 1 with RPC | Level 2 | |
| BlackSky | | | |
| EROS-B | Level 1A | Level 1B | |
| EROS-C | Level 1A | Level 1B | |
| GEOSAT 1 | | | |
| GEOSAT 2 | | | |
| KOMPSAT-3 | Level 1R with RPC | Level 1O with RPC | Level 1G with RPC |
| KOMPSAT-3A | Level 1R with RPC | Level 1O with RPC | Level 1G with RPC |
| Near Space Labs | | | |
| Pléiades | with RPC | | |
| Pléiades Neo | with RPC | | |
| Satellogic | | | Level 1 Level 3 |
| SPOT | with RPC | | |
| TripleSat | Level 1 with RPC | Level 2 | |
| Vision-1 | | | |
At least one geometric processing level per collection is always available on the console, but the availability of other levels may differ depending on the collection. Some levels are only available upon request.
| Collection | Primary | Georectified | Orthorectified |
|---|---|---|---|
| Beijing-3A | Level 1 with RPC | Level 2A* | Level 4* Level 4A* |
| EROS-B | Level 0A Level 1A | Level 1B | |
| EROS-C | Level 0A Level 1A | Level 1B | |
| Hexagon Aerial | | | |
| Landsat 8 | | | Level 2SP |
| Pléiades | * with RPC | * | |
| Pléiades Neo | * with RPC | * | |
| Sentinel-2 | | | Level 2A |
| SPOT | * with RPC | * | |
| TripleSat | | Level 2 | |
| Vexcel Aerial | | Oblique | Ortho True Ortho |
Rational polynomial coefficients (RPCs) are parameters used to model the relationship between the pixels within the imagery and their corresponding geographic coordinates on the Earth’s surface. It takes into account the various geometric distortions, sensor characteristics, and Earth’s curvature effects to allow accurate transformation between image and geographic coordinates.
An RPC file is typically delivered alongside imagery of the primary processing level and can be used for subsequent processing steps like generating DEMs.