The combination of GPS information and compass heading in the EXIF of pictures is quite mighty. Here an other idea what can be done with this information: With position + image direction it is possible to derive information on objects that are to be expected on the picture. The focal length and the image direction define a angle of view. Known POIs in the angle of view can be proposed to the user for tagging. The user simply checks the POI information he wants to use for tagging. EXAMPLE: You take a picture in London. The program finds POIS in the angel of view. You choose "Trafalgar Square" and "Nelson Column". The information is written into the IPTC. A proof of concept has been done by students! Find a detailed explanation with screen-shot here: http://www.gps-camera.eu/news/46-wissen/271-wwwbildschildde-aus-koordinaten-und-blickrichtung-auf-das-motiv-schliessen.html (German, use Google translate in the top right)) Contact me if you need test pictures with GPS + heading information! (http://www.gps-camera.eu/kontakt-impressum.html) Cheers Seb Reproducible: Always
Compass feature is already reported in bug #303062 Gilles Caulier
*** Bug 220846 has been marked as a duplicate of this bug. ***
Hi Seb, The auto-tags assignment feature based on IA introduced in next 8.3.0 do not solve this issue ? The deep learning neural network will analyze image contents and recognize subjects of photography in goal to tag item in the databse accordignly following a taxonomy logic. All is explained in details to the student blog entry here : https://community.kde.org/GSoc/2023/StatusReports/QuocHungTran#Add_Automatic_Tags_Assignment_Tools_and_Improve_Face_Recognition_Engine_for_digiKam Best regards Gilles Caulier