the current corrections mask (see below) and only pixels in the support of the corrections
mask have the scale factor applied to them. Furthermore, to avoid ‘divide-by-zero’ errors
and floating point overflow errors, an arbitrary lower limit is placed on all frame means: if
any frame has a mean pixel value that is less than 1.0e-10 it will be scaled by 1.0 and a
warning message will be printed to stdout (the console) if that happens.
The ‘
Apply dark field subtraction?
’ option, when selected, subtracts the loaded master
dark image from each frame captured. See the corrections algorithm in figure 6.5 for
detail. This option cannot operate unless a suitable master dark image is loaded.
‘Suitable’ means the master dark selected by the user must have the same dimensions as
the current image being captured and must have the same number of colour channels i.e.
a monochrome single channel master dark image if you are capturing monochrome
images and a 3-channel master dark image if you are capturing colour images. You chose
your what master dark image to load with the ‘
Select
’ button below this check box. This
brings up a file chooser dialogue box from which you can select a file. Only files in raw
doubles format (.dou with a .qih external header file) can be used as input. In the case of
loading a master dark field image for colour image correction, the file you chose must be
one of a triplet of files that have the same prefix name but end in ‘_R.dou’, ‘_G.dou’ and
‘_B.dou’ (or the same but with lower case ‘r’, ‘g’, ‘b’). Each raw double image must have a
corresponding external header file ending in ‘.qih’. You can load / choose any one of these
three and it will be accepted provided the other two members of the group exist and can
be read and are of the same dimensions to the one you chose. In the case of loading a
monochrome master dark field image the file name must end in ‘_Y.dou’ or ‘_I.dou’ (or
‘_y.dou’ or ‘_i.dou’) and have a corresponding ‘.qih’ external header file. Such images with
compatible file names for either colour or monochrome use can be generated by PARD
Capture using the ‘Save as raw doubles?’ option or using the ‘Intensity’ save-as format
described above.
The ‘
Apply flat field division?
’ option, when selected, divides each frame captured by a
pre-processed master flat field image. See the corrections algorithm in figure 6.5 for detail.
This option cannot operate unless a suitable master flat image is loaded. ‘Suitable’ means
the master flat selected by the user must have the same dimensions as the current image
being captured and must have the same number of colour channels i.e. a monochrome
single channel master flat image if you are capturing monochrome images and a colour or
3-channel master flat image if you are capturing colour images. You chose your what
master flat image to load with the ‘
Select
’ button below this check box. This brings up a
file chooser dialogue box from which you can select a file. File formats that are acceptable
for use are those with the following extensions: ‘.ppm’ (a PPM P6 formatted image),
‘,pgm’ (a PGM P5 formatted image), ‘.bmp’ (an MS Windows BMP formatted image of
either 8 bpp or 24 bpp) and ‘.dou’ (a raw doubles image with an accompanying ‘.qih’
external header file). When using raw doubles images for colour image correction, the
same rules apply as for loading a colour master dark field image (see above).
I mentioned that the master flat image is pre-processed prior to use for divisional
correction. What this means is that the image you select and load, automatically
undergoes the following processing prior to being finally accepted for use in flat field
correction (if it fails this processing you will get an error message telling you that the file
cannot be used for flat field processing). In the following description, all processing is only
done for pixels within the support of the currently active corrections mask:
OptArc AF51 Camera Page 81 of 99 User Guide v1.02
