Hello,
While working with images taken by the Elios drone, I noticed strange chromatic aberrations along some white edges.
- This halo seems to appear just on one side of the edge!
- It can happen anywhere in the image, not just near the borders.
- It doesn't appear on every edge even if they are equally white.
- It can be as large as 5-6 pixel wide
Looking at the attached images, I hope people can help me diagnose the exact cause of this and possibly suggest image processing improvement.
Regards,
Bruno Martin
Theoretically, Zernike polynomials (3rd order ) is used for aberrations . Curve fitting can be used for corrections ...I can help but need time !!
Hello Murthy,
Do you have some references on the kind of aberration I'm experiencing? The drone company suggested "Purple fringing" but that doesn't quite fits this kind of aberration when I looked for example of that.
https://en.wikipedia.org/wiki/Purple_fringing
https://en.wikipedia.org/wiki/Chromatic_aberration
As my goal is to have images to reconstruct the 3D scene by photogrammetry, I want to modify them as little as possible. The best solution may be to find a better apochromatic lens to replace the cheap one and get rid of the problem at the source.
The last thing I want would be to start working on a solution to the wrong problem...
Tech support from the company "explained" that it was caused by chromatic aberration and that the effect should get stronger the farther we get from the image center...
Somehow that explanation doesn't feel right as I have seen picture where this blue halo is very close to the center of the image (see third picture attached). Also, it doesn't address the fact that it always appears only on one side of the white grid, never on both sides.
Any better explanations someone?
Dear Bruno,
regardless the explanation, I think an easy and natural way to reduce the chromatic abberation before or after image acquisition would be optical or digital color filtering. Thus, you could place a color filter in front of the camera, or use only one color channel (for example the red channel) of the digital photo.
Kind regards,
Sascha Grusche
Well Mr Martin !!
If you do not like Optical math then the next best thing would be "just noticeable difference" (JND) taken from CIE color space representation of colors. Purple falls only on the x-axis (between RED-BLUE) , by inspecting the neighborhood and checking for continuity (or other metrics ) you can reduce the fringing.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images