A Fortran program (levela.f-used in the pipeline demo) was written that:
Read in test data
Read in the suite of 8 images per channel per exposure
(calibrated image, photometric uncertainty image, bad pixel
map, quality map, cosmic ray map, latent image, latent image
mask, ghost image).
Frame Quality Checks
1) Do statistics on the calibrated image
calculate min & max, min & max of good pixels (using badpix, qual,
cr masks to exclude pixels), mean of good pixels, means of several
regions surrounding the center of the image.
2) Count the number of pixels with various values (BDF=0,1,2,4,8,16,32,64,128)
in the bad pixel map.
3) Count the number of pixels (DQF=0,1,2) in the data quality image.
4) Count the number of pixels (CRF=0,1) in the cosmic ray mask.
5) Count the number of pixels (LIF=0,1,2) in the latent image mask.
6) Add information to fits header.
Is AOR executed OK?
Compare GLIMPSE AOR with AOR-related information in the calibrated image
fits header. Used GLIMPSE AOR
l303.12891-00.89655.20030306a.aor
to populate the fits header with some information. If the values in the fits
header don't agree with the AOR, ALERT comments are added to the fits header.
In this case, the ALERTS were forced to test the software and illustrate
the fits headers.
Verify SSC Processing Steps Done
Check that the SSC processing steps given in the header are "COMPLETE".
If not, write an ALERT comment to the fits header. In this case, the ALERT
comments were forced to test the software and illustrate the fits headers.
Note: A new calibrated image is created with additional fits headers.
The 7 ancillary images have not been changed.