Has anyone used software to automatically detect and rotate scanned photos to the proper orientation?
I have a directory tree of roughly 10,000 scanned photos from before 2002. Due to the number of photos, I didn't take the time to properly orient them in the scanner the right way. As a result, a large majority of the photos are upside down, sideways, etc. Obviously, there are plenty of tools to rotate photos in batch mode (I'm partial to IrfanView which can do it lossless), but I would have to manually sort them or rely on EXIF orientation data (which doesn't exist since they were scanned). From what I have read, AI software has progressed to the point that machine learning algorithms can detect which way the image should be oriented. I'm hoping to find an existing software solution to scan all 10,000 photos and determine the proper orientation. The result could put them in sorted directories, add metadata, or otherwise mark each file with the rotation needed. Even better, it could automatically rotate them the correct way.
I've read that Google Photos and Lightroom can do this, but I think it needs EXIF data. I've also read about the OpenCV image processing library that can use machine learning to do this, but I have no idea how to use it since I'm not a competent programmer.
Surely this is a common problem that others have solved? Has anyone found a solution do this? If not, any suggestions for the easiest/fastest way to achieve the desired result?
Thanks for your help!