Flatfielding

Note that because this process uses flatfields, you need to use a PCOT camera file which contains flatfield data! You can download the appropriate file for the geology filter wheel in the training model (TRAINING_GEOLOGY) from the camera definition files. It's quite big!

Download graph: flatfield.pcot

Figure: Flatfielding graph. Click on image to expand.

Flatfielding - flat-field correction - mitigates irregularities in captured images due to pixel-to-pixel differences in the camera's detector and optics. We use images of a flat white field (hence the name), which are stored in the cameras' data files.

The process performed by the graph above is as follows:

• input(0) - Bring the image into the graph from input 0. The input node must set the camera to a camera which has flatfield data.
• getflats(a) - this expr node looks at the filters used in the image and generates an image made up of the flatfields for those filters, with the same filter-to-band assignment.
• a/mean(a) - the flatfields are divided by their means on a band-by-band basis
• a/b - the image is divided by the flatfields.

This might seem a bit inefficient - why not divide the flats by their means when we create them using pcot gencam? - and indeed it is a bit slow. However, the process is going to change later when we start to use darkfields and other corrections.

As noted above, it's important to specify the correct camera when you capture the image and to make sure it's being loaded correctly. If you're loading raw images you will need to set up the raw loader in the multifile input dialog.

Figure: Flatfielding input setup, showing the camera selection and raw loader settings. Click on image to expand.

Presets for most settings can be loaded and saved using the "presets" box - see the multifile input docs for full details.

Results

To see the effects, look at the image in the input node's canvas. There are some white test patches. If we look at the third patch from the right, an artifact is clearly visible:

Figure: Portion of uncorrected image (normalised). Click on image to expand.

This artifact no longer appears in the corrected image:

Figure: Portion of corrected image (normalised). Click on image to expand.

You should also bear in mind that the overall intensity of the corrected image will be different, as we can see from spectra plotted from the middle of the largest white patch:

Figure: Spectra of uncorrected and corrected white patch. Click on image to expand.