Visualizing image deformation (object shrinkage)

I have two images of the same, complex object taken at two different times. The object has shrunk between images and I need to somehow represent this shrinkage in the form of a static 2D graphic. The shrinkage is quite obvious when "flickering" between the two images (appropriate for, say, a presentation) but I need to convey this shrinkage in a static figure for a paper.

Ideally, I would like to show some sort of color vector field (perhaps from quiver) superimposed on top of the final image. However, I'm struggling to extract the displacement field from these two objects. I first tried taking the difference between the two images and then taking the x and y gradients, using that as input into the quiver function. But this doesn't seem to capture the object features very faithfully, and it fails for larger deformations. I then tried computing a nonrigid deformation map using the registrationEstimator app. But this also has a hard time finding the relevant features and, consequently, the registered image doesn't resemble the reference image all that well.

Does anyone have some tips for how I could make some progress with this? Incidentally, it would probably be enough to map the displacement of the object edges (the interior has relatively little contrast), so this could potentially be combined with some edge finding routine, if anyone has a good suggestion.

Answer 1

[Martin Stolar 941141202]

If you are still looking.

PIV (particle image velocimatry)

It maps fluid flow as vector fiels but it might be helpfull.

https://www.mathworks.com/matlabcentral/fileexchange/27659-pivlab-particle-image-velocimetry-piv-tool-with-gui

[Simson Hutagalung]

How to generate deformation colormap 2D in matlab with data from excel?