Perceptual hashing

<p>Perceptual hashing is the use of a <a href="/facts/Fingerprint_(computing)/SSXY0pp2">fingerprinting algorithm</a> that produces a snippet,  <a href="/facts/Hash_function/rk6MlCt3">hash</a>, or <a href="/facts/Fingerprint_(computing)/SSXY0pp2">fingerprint</a> of various forms of <a href="/facts/Multimedia/PtJteq27">multimedia</a>. A perceptual hash is a type of <a href="/facts/Locality-sensitive_hash/hj8UllQG">locality-sensitive hash</a>, which is analogous if <a href="/facts/Feature_vector/nzAhYxfu">features</a> of the multimedia are similar. This is in contrast to <a href="/facts/Cryptographic_hash_function/cuxkgbST">cryptographic hashing</a>, which relies on the <a href="/facts/Avalanche_effect/hpYX5zu5">avalanche effect</a> of a small change in input value creating a drastic change in output value. Perceptual hash functions are widely used in finding cases of online <a href="/facts/Copyright_infringement/33JDsS5S">copyright infringement</a> as well as in <a href="/facts/Digital_forensics/MPF5Vdpq">digital forensics</a> because of the ability to have a correlation between hashes so similar data can be found (for instance with a differing <a href="/facts/Digital_watermark/73WXq5PK">watermark</a>).
</p>

Perceptual hashing open-in-new

Perceptual hashing