Normalization of images within series - Image content stabilization if image content has X/Y offset

[porg]

What I'm proposing for images in a series is called image stabilization in video production.

I often create a series of screenshots or mockups.
Despite good care, quite commonly in some images of the series I moved the image frame and/or the objects of interest, a few pixels on the X-axis and/or Y-axis. And if you go through the series in a presentation you then have jittering which is quite disturbing.

It would be great if there is a feature which can do image stabilization in such an image series and lets you decide the normalization method:
1) crop the images
2) add margin to images
3) leave image dimensions unchanged and only move image content on top of a transparent background or a monotone background color (if same on all 4 edges).

And do all this:

I) fully automatic (maybe possible with Apple's latest visual AI functions? or some rudimentary image stabilization library (no zoom, no rotation, only pixel offset on x/y-axis).

II) Semi automatically
• You step through the image series
• Per each image you have these choices:
a) Skip image (it is ok as-is)
b) With a crosshair mark the significant pixel (this is the reference point which exists in all images).
When done with the last image, and you have your method (1-3) chosen, GraphicConverter processes all files of the series.
After that you have a normalized image series (no more jittering of the image content).
c) Alternative workflow to B could be to extend existing functionality a bit:
c1) Keyboard shortcut to prev/next image
c2) Move content up/down (on top of transparent bg-pixels) or on solid bg-pixels with Add/Remove Border (-1 on top and +1 on bottom plus fill in bg-color to move image content 1 pixel up).
c3) Quick method to overlay prev/next image as your reference.
c4) Or possibility to continue going to prev/next images but keep the unsaved changes in all images in memory, and only at the very end, when everything is ok, persist the changes to all images.

[porg]

@forum_adm what do you think about this feature? I did not get a reaction back then. Hoping for it now. Thanks.

[forum_adm]

That is very special. The question is whether more customers need such a feature.

[porg]

I agree that this is a niche topic.
I understand that it receives no further action so far.
At least we have it documented.

1) For videos image stabilization has mainstream use cases undoubted.

2) For photos not really, because "hey, use a tripod" if you want comparable results photo to photo.

a) Maybe there is the demand to normalize a hand shot series, and by series I mean not shutter-down and 10pics/sec, but I mean snaps over the course of 1min or so, where the photographer slightly moves/jitters between.

b) But again, here my describes techniques would most likely not suffice. This likely will need rather an AI approach (comparative perspective analysis). And this is certainly out of scope for GC.

3) For screenshot series normalization there is likely really very little demand.

a) Amateurs mostly do 1 screenshot. Pros maybe do series. But pros should make sure to create homogeneous screenshots in the first place.

b) Only potential demand would be normalization of screenshots of the same software from numerous users.
Which may use different viewports, but where positioning the top-left corner of a window homogenously in all images would help comparisons (quick jumping through image series of different users).

End of documentation