The technology developed by Advestigo uses different techniques to extract content information from different media types, or Elementary Media - Text, Image, Sound.

Graphics and video content is fingerprinted on the basis of both detailed (small-scale) and global properties, to take account of the full range of information contained in the image. Two images can thus be compared at multiple different levels of detail. An important feature of Advestigo's technology is that it looks for similarities between compared images, and not at differences. The objective is not to say, for example, that a monochrome image is "more" or "less" like the original than a copy of half the original; Advestigo's fingerprinting demonstrates, for both cases, that they are copies of the same original.

• Fingerprint generation adapts itself to the content :
  • Automatic detection of elemental characteristics, or features:
    • Geometric patterns in graphics or video content,
    • Time interval in audio content,
    • Groups of concepts in text content.
  • Adaptive detection allows tracking these features through the document life cycle, independantly of modifications to the document.

• Calculating the characteristic properties of features :
  • Reduces sensitivity to noise,
  • Reduces sensitivity to normal life-cycle modifications or changes in the source (image colour changes, audio sampling, lossy compression...)
  • Comparing the characteristic properties of features allows the generation of a confidence ratio, based on whether the "suspect" file contains features equivalent to those identified in the reference or source data.
  • A heuristic approach can then be used to determine whether an image containing a certain number of identified features can be considered to be a copy or plagiarism of the original.