A secure digital watermark system for secondary distribution

Conventional digital watermark systems assume that the distributor (server) is a completely reliable organization. If the server is not a reliable organization, in a conventional system it is possible that suspicion may fall on a user who is not illegally distributing. Several methods have been proposed to handle this problem. In these methods, however, the system is essentially composed of two kinds of entities: the agent and the users. There is no discussion of a system composed of three entities - authors, agent, and users - where secondary distribution takes place from the author to an agent and then to the users. This paper proposes a digital watermark system that can prevent illegal conduct in a secondary distribution system composed of three types of entities - authors, agent, and users.