From this article http://people.csail.mit.edu/mrub/papers/vidmag.pdf I became extremely interested in the subject, but I couldn't find too much content that starts from the "basics". This process basically get a video, that has some translational motion, as input (or temporal motion, like colors, but my focus is on spatial motion), and by linear approximations of the movement it is possible to magnify it.
Where I(x,t) is the intensity of a gray-scale image in the point x and time t.
This part I could get it, and I was able to implement too with a simple function, but the clain is that with some process like spatial decomposition, temporal processing/temporal filtering, and some other things I was able to do this with an image, by finding it's B(x,t) and scaling it. This part is where I get lost, so if anyone knows about this process of eulerian video magnification and has any tips on what do I need to study, that would be of great help, or even if just knows some of these terms.
That is the process explained in the article:
"Overview of the Eulerian video magnification framework. The system first decomposes the input video sequence into different spatial frequency bands, and applies the same temporal filter to all bands. The filtered spatial bands are then amplified by a given factor α, added back to the original signal, and collapsed to generate the output video."
This article is great too https://dl.acm.org/doi/pdf/10.1145/3015573