I built a small Python project called Twistnshrink because I got tired of wrestling with the same annoying image problem.
The job sounded simple at first: take a pile of photos and merge them into one document. In practice, the images came from different devices, different orientations, and different file sizes, so the cleanup work became the real task.
On top of that, the final document could not get too large, which meant resizing and compressing everything before assembly.
The most useful part is batch processing. Instead of fixing images one by one, I can point the tool at a folder and let it normalize the whole set in one run.
Twistnshrink was built for exactly this kind of cleanup. It straightens, resizes, and compresses images so the final .doc or .pdf stays manageable instead of turning into an oversized mess.
The project itself was also a good excuse to tighten up the development side. I used AI-assisted tools while building it, set up proper version control, and added a GitHub Actions release pipeline so tagged versions can ship cleanly.
If you deal with the same kind of image cleanup, Twistnshrink may save you a lot of repetitive fiddling. GitHub: https://github.com/bilbilivo/twistnshrink