Legacy ImageMagick Discussions Archive

Thanks fwm42. I will have a look again. I have tried many of your scripts.

That makes sense. But looking at the magnitude of the slopes in your supplied image, I can't see any changes in brightness.

This is a crop from the combined sRGB image. The isolated bright red green and blue pixels are artefacts. They don't belong there - ignore chromatic aberration around the stars. I can perform the operation (whatever that is) on the colour images during stacking rather than the grayscale channels.

crop

I'm looking for a way of implementing a Winsorized Sigma Clipping routine in IM and perhaps the thread has steered away from that to finding hot and cold pixels by other means.

So, getting back on track and referring to the link is there a way to do this in IM or perhaps something similar. In a nutshell replacing outliers with the most valid neighbour value.

By "nearest valid neighbor in the sample", does that mean in spatial distance or graylevel? I assume it is nearest graylevel, is that correct?

Imagemagick does not have this function. However, I suspect it could be scripted (but likely would be slow) or someone might code it as another option to the Imagemagick -statistic function.

I think gray level is correct.

Winsorizing would be very handy.

The high-end program for this is Pixinsight
https://pixinsight.com

For free you can try
deepskystacker.free.fr/
Detect and clean remaining hot/cold pixels:
http://deepskystacker.free.fr/english/u ... osmetictab