Skyglow Project

[Babbalabba]

Hello experts, I am doing "Using a Digital Camera to Measure Skyglow" project.

I got a copy of the flanders article, and since I wanted to convert my measurements to magnitude per square arcsecond, I am following his steps. So far, so good, but I also want to make a full sky brightness map like he did. After taking all the neccessary pictures, is there a program that can stich the photos I took together (no fisheye lens/metal sphere ;S) and color code them based on their pixel values? I want to prove that globe lights that shine in all directions scatter more light into the sky than a full-cutoff light, which focuses the light down on the ground. I picked 4 sites, ranging from severe skyglow to moderate. I noticed that the sites with more skyglow had more lights, and the lights were shining very brightly and were not focused with a cone, or other material. This is why I want to make the full sky maps, so I can indeed see that different kinds of lighting do make a differnce in skyglow!

Thanks,
An eager student

[hhemken]

Hi Babbalabba,

I have successfully used Hugin (http://hugin.sourceforge.net/) a few times to stitch several photos together. For miscellaneous image processing, I would recommend GIMP (http://www.gimp.org/). Both are free, open source, multi-platform programs.

Good luck!
.

[theborg]

Hello Babbalabba,

Thank you for posting your question to Science Buddies. After doing some research concerning your topic, I think I've come up with something you may like to try. I haven't attempted it myself (my digital camera is not up for the task), but theoretically it should work. First though, unfortunately I don't see that there is any single software (at least in the range of free to cheep) that will stitch and perform individual pixel analysis and color code them based on intensity.

That being said, a good (free) image stitch program is called Hugin. A link to it's operation and the software can be found here: http://lifehacker.com/378490/stitch-pho ... e-software

I suggest taking your images encompassing the entire horizon and working your way up to Zenith with a nominal 20% - 30% image overlap, as suggested in the link. Using Hugin, stitch them together into a single image 0 to 360 degrees azimuth by 0 to 90 degrees elevation. Be sure to document very well which image corresponds to which look angle. Then using ImageJ, as suggested in the SB project procedures, analyze consecutive areas of the images using procedures from Step 7 thereby obtaining average pixel intensity for that area (vs intensity for every single pixel in each image). If you can, I would try to select sections of the images at some small increment like 1 or 0.5 degrees. Repeat this process until you've analyzed the entire panorama image you've created. Assign colors to intensity ranges (i.e. 0 < 100 = white; 101 < 200 = green; etc…) and plot them on a polar plot. You can do this in excel with some effort, but I’d look for a program, like matlab that can handle such operations much better. This will entail a lot of manual processing, but I think in this case “brute force” method is going to be your best bet.

Please post back with any other questions.


All that said, I’m not a computer programmer by trade, so if a programming expert has some other advice, feel free to correct or amend my advice.

[Babbalabba]

Thank you all SO MUCH for replying. I will definetly seek out the forums again for any other science help I need. YOu have made a happy 8th grader ;D.

I read your procedure for paranorama taking, and went to try it...alas. CLOUDS. The clouds continued noght after night, with no sign of let up, thereby rendering my process useless. Oh mother nature...

So I proceeded with data that was not skewed by clouds, only taking photos at the zenith. Sorry I let you down D; And myself, as I wanted a real challenge and the panorama idea seemed like one. But I am happy with the final result, and now I get to enjoy more time of my fall break! (yes, it's fall break here)

I definetly have something to keep me busy this winter break though.

Thanks again,
Emerson

[Babbalabba]

Thank you all SO MUCH for replying. I will definetly seek out the forums again for any other science help I need. YOu have made a happy 8th grader ;D.

I read your procedure for paranorama taking, and went to try it...alas. CLOUDS. The clouds continued noght after night, with no sign of let up, thereby rendering my process useless. Oh mother nature...

So I proceeded with data that was not skewed by clouds, only taking photos at the zenith. Sorry I let you down D; And myself, as I wanted a real challenge and the panorama idea seemed like one. But I am happy with the final result, and now I get to enjoy more time of my fall break! (yes, it's fall break here)

I definetly have something to keep me busy this winter break though.

Thanks again,
Emerson

[deleted-71771]

The experts provide an excellent approach to your attempt to consolidate pictures together. I would suggest that you make careful notes on how you have done this project, any issues that you faced, and lessons learned. It will be important that you be able to tell your story in a concise, easy-to-understand manner.