Image perceptual hash
A 64-bit fingerprint that stays close even after a re-export, crop or recompress. Useful for finding duplicates.
69 files analysed · 1.2 GiB processed
Upload
What is a perceptual hash
A perceptual hash is a short string computed from how an image looks, not from its bytes. Two visually similar photos produce two very similar hashes. Two unrelated photos produce hashes that differ in most bits.
You compare two hashes by counting bit positions where they differ (Hamming distance). Same photo, same hash. Same photo re-saved at lower quality, distance below 10. Different photo, distance well above 25.
What pHash is good for
Catching the same photo when it pops up on another platform after a re-export. Spotting bot-driven reposts. Building a dedup index for a photo library. Confirming that a stripped copy still shows the same picture as the original.
It is not a cryptographic hash. You can not reverse it to the original, and an adversary can engineer collisions, so do not lean on it for anything security-shaped.
What we compute
A 64-bit DCT-based hash (pHash) of the decoded image. We render it on the analysis page next to the SHA-256. Paste two hashes into your favourite Hamming-distance calculator to compare.