This is a tool I made for my friend. It basically takes in images/pdfs and arranges them into one clean, aligned file instead of dealing with a stack of separate scans. The purpose is to make stitching multiple scans easier.
It runs OpenCV.js compiled to WASM in the browser, handles deskewing scanned pages, lets you align pieces manually or automatically with control points, and exports the final stitched result as a TIFF file (the requested format from my friend, sorryyyyy).
I’m uploading the whole thing at once instead of logging it as I went, because I didn’t originally plan on putting it on Stardance. I built it as its own project, not with this challenge in mind. So this is not a live devlog, and I want to be upfront about that instead of pretending otherwise.
Same goes for the AI usage. I leaned on Claude a lot while building this, more than I probably would if I’d been treating it as a Stardance project from day one. I’m not going to dress it up as more hands-on than it was. It’s a real, working tool I understand end to end, built with heavy AI assistance, uploaded after the fact.