To identify audio and/or video content, related fingerprints must be computed of a content source (e.g. a movie, music video, or song) and sent to the fingerprint server for identification. This is done by the fingerprint Software Clients.

Ingestion fingerprint clients provide the following functionalities (see graphic here):

- Extract fingerprints from audio and/or video;
- Send fingerprints, including associated metadata such as title, assets ID, copyright holder etc. to the
  fingerprint server to be stored as ‘reference’ fingerprint in the hosted fingerprint database.

Identification fingerprint clients provide the following functionalities:

- Extract fingerprints from unknown audio and/or video
- Send fingerprints to the fingerprint server for identification

Displaying the identification result: this can vary from displaying only an aggregated business rule such as ‘take down/block’ or showing detailed identification results.


The software clients for fingerprint ingestion and identification allows for all levels of workflow integration: from simple scripting level to deep SDK integration (all software includes easy to use sample code, for ingestion and identification).

The software clients interface with the central fingerprint server, allowing direct ingest of new reference fingerprints and highly efficient mass identification of ‘candidate’ fingerprints.
Media/Format types such as, but not limited to the following codecs, are supported for ingest and identification: MPEG2, MPEG4, WMV, Quicktime, AVI, Divx, MPG, H264, Vc-1, mp3, AAC, and more.

The fingerprint registration and identification process is as follow:

Step 1: Creative works are fingerprinted from live broadcast streams or archived material and 
            registered in the Media Database.

Step 2: Business rules and metadata are assigned to each individual fingerprint.

Step 3: Uploaded content on a UGC website or P2P network that has installed the Mediahedge detector, 
            is fingerprinted and recognized.

Step 4: The system search for specified media content and fingerprints are generated from
            detected media content.

Step 5: The generated fingerprints are compared with fingerprints previously stored in the central 
            Media Database.

Step 6: If there is no match: no action is taken (content is allowed and published). 
            However, in case of positive identification or match, the pre-defined business rules are applied.