I have a 3D data set sampled on a regular grid shown in the Figure below. This data can be interpreted as an image where the point altitude is equivalent to the pixel intensity value.
In the middle of my 3D data, I have a peak, that becomes a bright spot when the data is interpreted as an image. My goal is to remove this hot spot from my data.
I was told to try some image inpainting techniques to remove it without introducing too many artifacts. After searching available implementations, I've found two algorithms already implemented in the opencv library:
- one is the Navier-Stokes Inpainting by Bertalmio et al.;
- the other one is a fast marching based approach by Telea;
but I don't know which one to use.
Is any of these methods suitable for my task? Which of these methods (if any) should I use?