Automated QA
Umbra performs the following automated QA checks on each Collect* before delivery:
- Blank: Blank Image Detection attempts to identify images with fundamentally no structural content - an image product dominated by noise.
- Eclipsing: Eclipsing is receiving data from one transmit pulse while busy transmitting another pulse (loss of data). In a black-and-white representation of a Collect image, this often looks like a white-to-black gradient across the range direction of the image.
- Mispointing: Mispointing is errors related to the Satellite not pointing directly at the center of the collection target. Mispointing also often looks like a white-to-black gradient across the image, but the gradient may be in any direction.
If a Collect fails any of the above metrics, the Collect status will terminate with CORRUPT instead of DELIVERED. The associated Task will terminate with status ANOMALY instead of DELIVERED.
Data products associated with CORRUPT Collects that fail QA will not be visible in Canopy. Data products for these Collects will also not be transferred to customer platforms via DeliveryConfig.
Exceptions
Extended Scene Size (Spotlight footprints other than 4x4 km) undergo a restricted set of QA checks subject only to mispointing and eclipsing.
Other QA Checks
Umbra performs additional automated QA checks that include but may not be limited to:
- Geolocation Calibration: This is averaged calibration object location error, in meters. This check is only performed when specific calibration sites are collected.
- Radio Frequency Interference (RFI): RFI occurs when a transmitter(s) produce radio frequency energy that interferes with the satellite mission of collecting data in support of image formation.
These additional checks are used to ensure normal system operation. They are not used in determining whether a specific Collect passes or fails QA.
Updated about 1 year ago