PDQ

Open perceptual hash for images, with a matcher to go with it.

Made by MetaHash matching and databases

Visit github.comRepository

About PDQ

PDQ is Meta's open perceptual image hash, released in 2019 as an open alternative to PhotoDNA. It produces a 256 bit hash that stays similar for visually similar images, so you compare hashes by Hamming distance and set a threshold. Implementations exist in C++, Python, PHP, Java and Rust, and a FAISS based matcher handles millions of hashes.

It's the format GIFCT and many newer lists use, and because it's fully open you can inspect and run it anywhere.

Reviews

Be the first to review it

Anonymous
Would you use it again?

Signed as your field and level, never your name.

Know a team weighing PDQ? Send them this page.