This release brings significant improvements to the automatic document detection. The new version of the segmentation model performs better on magazine-style or textbook pages. The app can now handle documents with missing corners — for instance, when pages are folded or stapled.
Other changes:
- New Czech translation, thanks to @rr-vesp
- Removed duplicate libraries from About screen, thanks to @manish222261-code
Certificate SHA-256 digest: 820710961328e1834bcb8158092e6b6c5b519fcd3bdb2e5319ebdaf08f9ea1e8