DSB:
> It would be (relatively) quick to modify the existing astResample methods
> to have an option to multiply the output pixel values by their area
> (specified as a number of input pixels), but this is not a very accurate
> way of conserving flux.
Agreed. That's why I wrote "even".
> A more accurate but slower way is to create an astRebin method (as an
> alternative to astResample) which passes through the input image pixels,
> placing each one into the output image at the appropriate position.
One of the ADASS talks (Montage?) emphasized that fidelity was important
first and that they had achieved conservation of flux to 0.1% in their
mosaics, but that it was quite slow because of all the spherical trig.
required to piece together the flux contributions.
> cater for input pixels which project into more than one output pixel, it
> would be possible to sub-divide each input pixel into smaller sub-pixels.
> This could be done recursively, using an algorithm something like:
>
> divide the current (sub)pixel value by 2**ndim.
This assumes a constant flux across the (sub-)pixel. While it preserves
the flux, it may affect the object's profile. You may need to worry
about the rates of change across the (sub-)pixel for undersampled data.
There's an article about this in an ASP Conference Series on Precision
Photometry (vol 189), or the earlier ASP Conf. on CCD Reduction (vol
23), probably by Steve Howell.
Malcolm
|