I'm sitting on approximately 1000 scanned photos. Most are about 20-30yrs old so in pretty good condition. A couple are much older.
These are the issues that I would like to fix in order of prevalence in the set:
- Photo paper texture makes it look like it is a scan vs a photo
- Adjustments for brightness/contrast/color etc
- Photos are blurry or out of focus
- Photo is damaged (torn, wrinkled, etc)
- Photo is really old
I'd like to batch process as much as possible. Not exactly sure how to do #1, #2 can be done by alot of software with their "auto adjust" settings. I'm thinking maybe Irfanview? or Photoshop? Not sure what other tools there are.
There are some machine learning or neural networks models out there that can fix #3-5 and maybe #1-2. However many are focus on the face and some go way too far and basically replace the face so it looks super fake. Just looking through gitlab there are alot of option. Then there are all the random "AI" websites. Is there a comparison somewhere?
After I'm done batch processing, is there an easy way to compare original with new photo? The scans are all TIF files. I figure I'll save the edited ones as PNG with the same base name.
I'd appreciate any recommendations.