[AstroPy] An Odd Result on an All-Sky Image--Equal Images and Bars?

Jim Vickroy Jim.Vickroy at noaa.gov
Wed Apr 15 17:48:18 EDT 2009

Wayne Watson wrote:
> (Well, I guess the moderator didn't release with two 1M fits files. 
> I'm posting it without them. Jumping down to HERE might be best. The 
> first part of this discusses the two files, but maybe it strikes a 
> cord with some. The most important part might be about comparing two 
> files. I'm attaching a small file of the barred image I discuss. I 
> think if I keep it down to 40K, it'll pass through to the list. )
> With help from posts above, I finally got a fits image from some old
> data I was using for test purposes.  As displayed by the sentinel
> software, the frame was dark with  the moon in the center a as small
> bright circle. I used a program, ccdsoft*, which displays fits images.
> It had much greater contrast. One could see the circular horizon, and
> trees along the horizon, even my house in the lower right. Neighbor's
> lights were visible in the lower left. That's not really anything new,
> but why the difference in the two images as displayed, I asked myself? I
> tried the image in DS9, and it showed it to be almost identical to the
> image displayed by sentinel.
> It's quite possible that ccdsoft has some auto contrast feature that was
> applied, but I doubt it.  I saved the odd ccdsoft displayed image to a
> fits file. The file name ends with ...x.fits. I'm pretty sure the pixel
> values will be different from the other file. Is there a way to compare
> the two easily? This is just curiosity.
If I remember, you know how to use pyfits to load FITS files, so load 
the 2 files and compare (as desired) the (primary) data units (DUs) of 
each file.  Since pyfits returns the DUs as numpy arrays (ndarray), the 
full power of numpy is at your disposal for the comparison.
> Perhaps more importantly here, and this I had no immediate plans of
> discussing here, since it was currently less important to putting in a
> "save to fits" ability; how does one get rid of those vertical lines
> (bars)? They are from the video camera working at high gain, I'm pretty
> sure.  
I did not see the vertical lines in the attached jpeg, but you may want 
to take a look at scipy.interpolate.Rbf 
and or the topic of image */inpainting/* in general.

-- jv
> Dark frame subtraction doesn't help. Although removing the lines
> is interesting here, it gets to be important in the situation where I've
> stacked 256 images of the sky (which does show many stars--and noise).
> BTW, the unstacked image is a  1/30th sec exposure. The purpose there is
> to bring out dim stars. It works fine, but those lines probably won't
> help star identification. I would think if it can be done with this
> image, it could apply to the stacked images. Comments?
> I'm attaching two fits files. Note that in the fitss file provided that
> something of a coordinate system, 3 lines, is shown. That's pure
> experimentation from some other interests.
> Interestingly, in DS9, the image is upside down (neither file is the
> display screen of the DS9).
> * BTW, this is the first time these images have been available to
> ccdsoft or ds9. ccdsoft does have IP capabilites, and sentinel has none.
> As a consequence, they are seldom seen as provided by ccdsoft or any IP
> program. Just playing around with a few IP controls on ccdsoft was
> interesting.  That's a real plus having IP.
> ------------------------------------------------------------------------
> ------------------------------------------------------------------------
> _______________________________________________
> AstroPy mailing list
> AstroPy at scipy.org
> http://mail.scipy.org/mailman/listinfo/astropy

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/astropy/attachments/20090415/f4e85fe3/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 27746 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/astropy/attachments/20090415/f4e85fe3/attachment.jpe>

More information about the AstroPy mailing list