2

I have a large 2000 frame series of night sky images taken from a gopro hero3. Which is in itself silly because GoPros have bad low light performance. I've done some light math on the images, and identified star position "arcs" over the 2 hour period.

Is there a way, using the star movement, to solve for the distortion of the lens AND the star rotation? Perhaps using OpenCV, or barrel distortion correction formulas?

I feel like there should be a generic way to solve the pixel offset equation, regardless of the interaction between the lens distortion and star rotation. Any ideas? I believe I have all the info needed, about 30k data points across 250 stars of: (star_id, time, x, y) with good coverage, see the fast graph image below.

star tracks

Benjamin H
  • 5,164
  • 6
  • 34
  • 42
  • 1
    This isn't really about programming but about image manipulation algorithms. This might be better suited on [dsp.se]. – Jim Garrison Feb 22 '14 at 00:05
  • I was unaware of that forum - thank you very much for the pointer! – Benjamin H Feb 23 '14 at 05:47
  • Question moved to [appropriate forum on DSP StackExchange](http://dsp.stackexchange.com/questions/14602/motion-and-distortion-estimation-using-multiple-still-images) – Benjamin H Feb 23 '14 at 06:09

2 Answers2

1

To compute distortion parameter you need to have a certain scenario. Just having enough points doesn't mean they are useful. In your case it would even require additional math work to compute the time corrected object coordinates.

On the other hand you don't really need to use your dataset to compute the distortion parameter (DP). With this camera you would probably not achieve an accuracy better than ~ 1/20 pixel because it is simply not meant to measure.

What I suggest is to create a good calibration scenario and calibrate the the camera where you compute the focal length, offset (x0,y0) and the at least three barrel distortion parameter from the cubic function A1,A2,A3. Afterwards you can apply the barrel parameter, to your images.

Be aware of that a calibration scenario yields (always) local DP parameter, which work best only for the current scene and not for your night scenario. But since the cam / lens combination is not made for precise measurements you won't really get better results when you compute the DP from your night scene.

8472
  • 151
  • 1
  • 8
0

If you need just camera distortion correcting (and don't want to do it by using another difficult methods because "science is beautiful") you can use camera calibration from openCV.

Vit
  • 793
  • 4
  • 17