Content fingerprinting creates an identifier based on the visual content of an image or video, rather than the exact bytes of the file. Unlike a simple file hash, which changes completely if a single byte is altered, a fingerprint is designed to stay similar when content is compressed, resized, or lightly edited.
This is different from watermarking. A watermark is something added to the content itself; a fingerprint is calculated from the content and stored separately. That means fingerprinting can help recognise your content even if it was never watermarked. Heavily edited, cropped, or re-recorded copies can still be hard to match.
Privly uses content fingerprinting alongside username search and reverse image search to find copies of your content online, so takedowns can be filed against the matches.