Multimedia Authentication

This use case discusses a company which is responsible for the implementation of a network of security cameras in a factory. They are connected to the Internet and automatically archive all images to a central server. One day a box containing a new prototype is found missing from one of the production areas. The camera records show how trespassers have broken into the facility. They have also recorded the face of one of the workers. He claims to be innocent.

In order to be sure that the camera records are reliable evidence, the company needs means to verify that they have not been tampered with. This can be done by means of cryptographic approaches [1] but digital watermarking can offer the useful advantage of detecting which parts of the camera records have been modified. This is achieved by encoding a special signature called a fragile watermark into the records which becomes broken in the modified parts. Furthermore, there are specialized digital watermarking methods that allow the recovering of the original modified parts [2].

The term fragile watermark (in the context of digital watermarking) refers to a watermark which is broken if the host medium is modified.In this use case, digital watermarking can confirm that the area of the camera records showing the face of the worker has been tampered with. It may be even able to recover the face of the actual perpetrator. This leads to the conclusion that the worker is innocent. A other information regarding how the theft has been carried out is genuine and can be relied upon in the course of investigation.


References:
[1] J. Fridrich and M. Goljan, "Protection of Digital Images Using Self Embedding," in Symposium on Content Security and Data Hiding in Digital Media, New Jersey Institute of Technology, 1999.
[2] G. Friedman, "The Trustworthy Digital Camera: Restoring Credibility to the Photographic Image," IEEE Transactions on Consumer Electronics, vol. 39, no. 4, pp. 905-910, 1993.