Closed Destiny-Dream1 closed 12 months ago
Hi, this is very likely due to how we do the color correction. I don't really know how the image on the left was obtained, but it probably used a different correction. In particular, different choices of rgb2xyz matrices could lead different looks (https://github.com/Jamy-L/Handheld-Multi-Frame-Super-Resolution/blob/e9b9c00059ff591f2d6a65052c5ff34bf3c34ec7/handheld_super_resolution/raw2rgb.py#L115C9-L115C9). This not really my field of expertise, so I dont really know which matrix would be the "best".
Destiny-Dream1,
It looks like a white balance problem to me. You closed the issue so I suppose you solved the problem. If so, what was the cause so it can help other users? Thanks!
Thomas
Hello, I'm not entirely sure how I obtained this photo. Afterwards, I couldn't reproduce this result by converting DNG to any other file format. I personally think it might be an issue with RGB and BGR, so I've closed this issue for now. Thank you very much for the responses from the authors.
Ok thanks! If it's a matter of Bayer pattern, eg you have BGGR instead of the expected RGGB format, you can rotate by 180 degrees the raw images right after reading the raw files and you should be good.
Hello, I noticed that when running your program, the colors in the generated results appear different from the low-resolution images. I'm not sure if this is normal. I have tried adjusting various parameters, but I couldn't match the colors with the low-resolution photos.
I have attached my results. The left side is the low-resolution first frame, the middle is the super-resolution result with color correction, and the right side is the super-resolution result without color correction.
Thank you so much.