Solving a wide-field image

[Luca Rosignoli]

Hi, I'm trying to solve a full-frame images with a fov of about 60°x90°. Is it possible with local astrometry suite? If it is, what are the best configuration options?

Thanks.

Luca.

[Dustin Lang]

Hi,

It might work... depends on how extreme the distortion in the image is.  You'll want the "4100-series" index files http://data.astrometry.net/4100/

And you'll definitely want to tell solve-field the minimum size of the image, eg

solve-field --scale-low 45 --scale-units degwidth --downsample 2 --crpix-center ....

We do often find that very wide-field images don't work very well... you can try cropping out a central region of the image where the distortion isn't so bad.

cheers,

dustin

--
You received this message because you are subscribed to the Google Groups "astrometry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to astrometry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/e1d1fd04-061e-4d38-b0db-bd1f0b4083a4n%40googlegroups.com.

[Gaetano Zarcone]

Hi, I'm working also on wide field of view images (not 60x90 deg, but 15x10)

To solve this type of images I use :

 --tweak-order 5

--crpix-center

--downsample 2 

--use-sextractor

4100-series index files

and it works!!

Cheers

Gaetano

To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/CA%2BFzHBBcqji0aa4vjW1HwtzjpQSE3%3DQ9Yshq%2BFeAfSwNJPoTtw%40mail.gmail.com.

[Rosignoli, Luca]

I downloaded both the 4100 and 4200 index series. Is there a command to force the solve-field to use only the 4100 series?

Luca  

[Dustin Lang]

Hi,

If you set the image scale (--scale-low, --scale-units), then it will only load index files that are relevant, so for wide-angle it will choose the 4100-series.

cheers,

dustin

[Gaetano Zarcone]

You have to change the cfg file and from there force astrometry to use the 4100 index

You have to give the path where astrometry searches for the index.

 In cfg file you will find the line "#In which directories should we search for indeces?", and below this line write "add_path your_index_path"

Hope being useful

Cheers

[Gaetano Zarcone]

Or as Dustin said above 😅

[Rosignoli, Luca]

I would like clarification about the meaning of the scale low parameter. Should this parameter be set at the arcsec/pix scale of the image? 

Thanks,

Luca

[Dustin Lang]

Hi,

You can set the units it is in using the --scale-units parameter.  For example,

solve-field --scale-units arcsecperpix --scale-low SCALE_IN_ARCSEC_PER_PIX

or

solve-field --scale-units degwidth --scale-low SCALE_IN_DEGREES_WIDTH_OF_IMAGE

Either way, scale-low is a strict lower limit, so if you know your images are 60 degrees wide, set the limit to, say, 50 degrees, or 55 degrees.

cheers,

dustin

[Rosignoli, Luca]

Very clear, thanks.

Nevertheless, I've just followed your initial advice, and I start to divide the image in four sub-images of about 20°x30°. Unfortunately the image are also quite noisy, due to the low exposure time (less than 1 sec).

Are there some parameters to fine tune for noisy images?

Thanks for all the support.

Luca. 

[Dustin Lang]

Hmmm, usually "--downsample 2" helps; you can try "--downsample 4" also.  To force it to accept more (marginal) star detections, you can use "--nsigma 6" or 5 (default is 8).

cheers,

dustin

[Rosignoli, Luca]

Can I also ask another question? It covers another topic, so I have to open another question or can I ask here?

Luca.

To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/CA%2BFzHBB7R6it6Qeypgakjm%2BXW7e0C-A5tKjeODc187YeCH4a7Q%40mail.gmail.com.

[Dustin Lang]

Go ahead :)

[Rosignoli, Luca]

Well, let me briefly explain the contest.

I'm now working on a project related to Space surveillance and tracking, using wide-field (60°x80°) optical images. Now I start to build the processing pipeline, and of course one of the crucial (and also critical) steps is the astrometric solving of the image. I must to reach at least the 1/10 of pixel precision in the astrometric solution (about 5"-6" since the pixel scale is 57"/pix ). Initially, I try to solve the full image at once but nor the Astrometry and neither other solve like PixInisght solver script can solve the full field image with the precision required. In addition, the image are also quite noisy, due to the low exposure time required (0.1-0.05s of exposure).

So I think that the only way possible is to divide the image in sub image with smaller fov like 10X10 or 10x20 deg of fov, and I have to optimize the parameter in order to achieve the best precision possible in the astrometric solution.

By the way there is another important aspect, that is the topic of the question. This pipeline should process hundreds of images per night, and it should produce the results as fast as it can! 

So I was wondering, is it possible to parallelize the astrometry solver in order to compute multiple solutions at a time? And also, is it possible to allow the astrometry engine to use the GPU to speed up the process?

I hope I was clear, and I know that it is a HUGE question. I hope this question will inspire you more than scare you 😂

Luca.

[Dustin Lang]

Hi Luca,

Can you say a bit more about this 1/10th of a pixel requirement?  Is that a requirement on the pixel precision on each star, or on the absolute astrometric error on any point in the image, or ???  As you may know, you need a fairly bright star to get 1/10th of a pixel accuracy in measuring the pixel-space position of a single star, since positional accuracy goes like ~ S/N / PSF size, assuming well-sampled images.  With short exposures and a super-wide-field, that might be a challenge.

Astrometry.net was originally designed to deal with pretty narrow-angle images where the distortion due to different projections isn't a major issue, but for your super-wide-fields, it must be a big concern.  So it may be that you'll want to apply a distortion-correction to your pixel coordinates before running the astrometry.net code.  Or try using the --predistort option, after generating a good distortion solution.

The super-wide-field images you're processing should be fairly easy for the code to solve, because there just aren't very many stars at that brightness level.  The biggest thing you can do to speed up solve-field is to give it accurate --scale-low and --scale-high bounds, and make sure the star detection is correctly finding and localizing the brightest stars in the image.

Currently the solver is single-threaded.  If you're batch-solving many images, you could of course put a wrapper on it (eg, call it with xargs) to run many copies of the solve-field executable.  It would be possible to extend it to run multi-threaded, but that would be real work.

I don't think GPUs are a good match for this problem, because the main compute cost in astrometry.net is searching for nearby objects -- first, searching for nearby quadrangles of stars, then searching for nearby stars that should appear in the image.  These are stored in k-d tree data structures, so they require a large number of branches, and GPUs don't do well on that kind of problem.

cheers,

dustin

[Rosignoli, Luca]

Yes sorry Dustin, the 1/10 pixel precision on the coordinates regard the whole pixels in the image. Not just the stars.

Luca.

[Dustin Lang]

I haven't really measured astrometric accuracy that way before, but I think that's going to be an extremely difficult target to achieve.  Even if you had a large number of stars, mapping out the distortion field to that accuracy would be difficult.

cheers,

dustin