Astronomical source extraction in non uniform background

[sonutabitha]

I have been working with some Herschel images and I am finding it difficult to extract the astronomical sources from the image. The image background is extremely non uniform so i am not able to differentiate between the background and the source. I used some tasks in HIPE like source extractor but am sceptical about it's reliability as it assumes the background to be uniform. Is there any software to deal with my problem? Is there any way out?
Kindly help

 

[mfb]

My background is particle physics, not astronomy, but looking for structures in nonuniform background is a typical problem in particle physics as well.

Do you have some model to describe the nonuniform background? Non uniform in which way?
There are various fitting tools that look for peaks above a (relatively) smooth background.
If the background has structures that look the same as the actual sources and no theoretical description for them, then I don't see hope.

 

[laner]

What kind of sources are you attempting to extract? What Herschel instruments are you using?

If you are looking at star forming regions then getsources is the way to go for sure. It is a bit involved to get it going but it works like a dream. Since star forming regions always have emission on multiple size scales the algorithm focuses on separating compact sources from the irregular background with iterative convolutions. Even if you aren't working with maps of star forming regions give it a try. It's built for Herschel so it plays nice with the all the PACS and SPIRE wavebands.

http://www.herschel.fr/cea/gouldbelt/en/getsources/

If you aren't looking to get involved with a more complicated algorithm then you may have some trouble depending on what exactly you are observing.