Home > Technology
Theraography - Advestigo's patented content recognition technology - can be applied to a variety of mutlimedia formats to generate a "fingerprint" based on file content. Specific algorithms are used depending on the media type - audio, video, graphics, text, etc. - but with a common representation, simplifying fingerprint database management. Where files contain multiple different media types (for example audio+video, or text+graphics) each media type is analysed using the appropriate algorithms and multiple fingerprints are generated. Content fingerprints are independent of file types and encoding - a fingerprint generated from a JPEG image, for example, will recognise the same image in another graphics format.

In this way, a document is represented by one or more fingerprints which allow all or parts of the content to be traced even after substantial modifications - cut and paste into other documents, changes of format or coding, additions or deletions. Advestigo's technology looks for any single point of similarity in the content, rather than calculating a "similarity factor". A copy is defined by the existence of common characteristics, rather than the degree to which the original may have been altered. Because the technology relies on fingerprinting it can be applied to any digital content, including all pre-existing data. To identify if a suspect document is a copy of, or contains content copied from, a reference document, all that is needed is to generate the fingerprints for the original and the suspect and to compare the fingerprints. Advestigo's technology can also be used to generate fingerprint databases of original works, to simplify and accelerate tracking of the dissemination - legitimate or otherwise - of content held in online databases or libraries.

An innovative technology:
  • No "Watermarking"
    • Can be applied immediately to existing content
    • Does not alter or "pollute" existing data in any way
    • Cannot be masked or removed
    • Can recognise re-created content - retyping, loop recording, camcorder capture...
  • No Encapsulation
    • No need for specific applications or infrastructure, on server or client
    • Tracks content dissemination, whether legitimate or unintended, by authorised users
    • No impact on standard document life cycle
  • No Electronic Signature
    • Based on content, not on file