Cees Bassa
Cees Bassa
The FF frames are stored as 8bit FITS extensions. For the MAXPIXEL, MAXFRAME and AVEPIXEL extensions this is not an issue, but for the STDPIXEL frame the use of 8bits...
The raw observations, as they were submitted, should be downloadable, preferably through an API. The API could provide a filter for satellite ID, observer, observation format and a time range.
It would help observers if they could have access to the TruSat TLEs through a single, static URL. Observers will want to pull the latest TLEs of all objects from...
The Hough3D output sometimes gives `bz` values that are zero, giving a divide by zero error in `fourframe.py:591`
In some cases where an astrometrically solved pointing clouds over and no stars are detected, no updated WCS is written to the FITS file. Instead, write the predicted WCS.
Documentation is sorely lacking. Fix this! See Jebba's STVID clone at https://spacecruft.org/spacecruft/stvid
The current astrometric calibration strategy is to check FITS files for the first with enough stars before attempting calibration with `solve-field`, and to continue until the astrometry of a FITS...
Processing depends on 3rd party executables (`source-extractor`, `hough3dlines` and `solve-field`). Add in a check to see that these executables are available in the path, and check four different naming schemes...
SKimage has morphology functions that can remove blobs. ``` from skimage import morphology morphology.remove\_small\_objects(data, min\_size=32, out=data) ``` This approach seemed to work well on masks from `sigma` frames [`(max -...
Satellites which are predicted to be visible but are not detected as trails (or track & stack), should be logged to allow manual inspection.