One feature I’ve been missing in open source image software is face recognition. I’ve been using Picasa for tagging faces in my photos for years now (running in Linux with wine emulation), but I’d really like to switch to something more open.
In 2020 (with the 7.0.0 release) digikam seems to get closer to this. In the current 7.0.0 release however, I couldn’t get faces recognised for some reason. Ok, the big changes from Google’s Summer of Code are planned to be merged in the 7.2.0 – release date somewhen end of the year.
I’m really keen to exchange my Picasa solution with something nice!
Update: Just tested post 7.1.0 git version – good news: Face recognition works nicely.
I’m missing one thing however: I tagged all my pictures using Google’s Picasa. Digikam nicely reads and displays the tagged faces, however Picasa marks images in a different way (wider rectangle) which makes those faces unusable for digikam training.