I recently inherited responsibility for a website that received a notification that a user had uploaded a copyrighted image years ago and that it was not licensed.
The Wordpress-based website has thousands of images under wp-contents\uploads.
Are there any techniques to script scanning all of the images and flag just those images that I should further investigate to see if they're also unlicensed/copyrighted?
Not perfect by any means, but I tried to do my due diligence in identifying any other images that could possibly be copyrighted & not licensed for use so that a human could decide whether it goes or stays. (There are about 10,000+ images uploaded on the site). There are probably a ton of better ways to do this, but this is my solution.
My solution was to use the Get File MetaData powershell script to recurse through all of the files that I downloaded locally from the Wordpress site - links to the script & background info can be found here - https://devblogs.microsoft.com/scripting/use-powershell-to-find-metadata-from-photograph-files/
It took a long time for the script to recurse through all of the subdirectories & create an object that contained all of the metadata - close to an hour to finish & give me the command prompt back after creating the $picMetadata object in powershell. The command I used was:
$picMetadata = Get-FileMetaData -folder (gci C:\Temp\WordPress\wp-content\uploads -Recurse -Directory).FullName
Then I exported the Name, Copyright, Dimensions and Path fields to a CSV file so that I could bring up the results in Excel & sort by the Copyright column to see if any of them had values in that field.
$picMetadata | select Name, Copyright, Dimensions, Path | Export-Csv -Path c:\scripts\picMeta.csv -NotypeInformation
Out of the thousands of images, only a hundred or so had anything in the Copyright metadata field - most that did had public domain notices in them, about 50 had either photographer names in them or attribution instructions in them.