I have two images, one is as seen by the camera and the other is a simulation (and hence has less details, though the resolution is same). I want to mainly correct for any translation and scaling between the images. [enter image description here](https://i.sstatic.net/AVWEAp8J.png)

I tried common image registration methods based on cross-correlation, FFT and optical flow but none of them were good enough.

Now I want to do it manually in an interactive way. By simply dragging one image over the other and then noting down the required translation. And something similar for scaling. Is there any free software which can help me do this? My images are currently in numpy format.