[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

[APML] First results to the challenge



Hi guys!
 
It seems that you forgot the challenge... well, I hope that this will encourage you to participate
 
http://download.astrophoto.vze.com/masks.jpg
http://download.astrophoto.vze.com/results.jpg
Here are the proccessing steps:
 
All the masks, and threatment, was made using PixInsight.
 
The first mask was made using a speccially made tool for uneven field illumination corrections, named Dynamic Background Extraction (DBE). Basically, it fits a surface function to some sample points and generates a new image with the background only. It works in a "per channel" basis. I did this mask with about 70 points in all the image, avoiding the Milky Way's bright zones. The continuity order in the function was 2, and smooth of 0.02.
 
The second mask was made throught wavelets (A-Trous algorithm). I choosed a dyadic secuence, and deleted the firts 6 layers. To the 7th layer I applied a very aggressive noise reduction. If I'd deleted more layers the bright influence of the MW vanishes, but the illumination in the resulting image doesn't match with the original one as good as this one.

The last one was generated using median and gaussian blur filters. Three iterations of the hybrid way median filter (15px radius), and two gaussian filters (order 1/4 and 2, 20px std). Again, a larger filters generates a smoother image, but it doesn't match the original one.
 
 
As you see, the last two ones still has a considerable amount of the brightnes of the Milky Way. Perhaps with a more "dedicated" procedure it could be removed, but seeing as how much easiers DBE was, i did not tried to do that.
 
 
In results.jpg you can see what happens if you substract the mask to the original image, and make a levels adjustement (B&W clipping of 0.01%, and gamma changes to obtain a median of 40 per channel)
Also you can see the final histograms, at 8bits resolution (i worked at 32bits, and that is the reason I think  it is so smooth)
 
Hope you'll like the comparison... and that you'll try to beat DBE's results =)
 


Regards,

Carlos Milovic F.
-------------------------
Astro & Photo - CMF
http://www.astrophoto.vze.com
-------------------------
Visita "AstroFoto", el foro de astrofotografía en español
http://espanol.groups.yahoo.com/group/astrofoto



    Antivirus • Filtros antispam • 6 MB gratis
    ¿Todavía no tienes un correo inteligente?
_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo