Secure watermark detection schemes

In traditional watermark detection scheme, a prover exposed a watermark to be present in a digital data to the possible dishonest verifier. However, a potential attacker is able to destroy or remove the watermark entirely once secret information like the watermark or the embedding key is known. Some of previous schemes proposed as solution haven't achieved desirable result really. Their lacking of security and validity is the most serious problem. In this paper, we propose the secure watermark detection schemes for both blind watermark scheme and non-blind watermark scheme based on zero knowledge protocol and public-key encryption scheme in order to solve this problem. There no secret information that can be used for removing watermark is disclosed during the detection process. Both security and validity of watermark detection can be significantly improved using these two schemes.