Joint fingerprinting and decryption with noise-resistant for vector quantization images

With the popularity of the Internet and development of multimedia technology, media distribution and traitor tracing issues have become critical and urgent. In this paper, a joint fingerprinting and decryption (JFD) scheme based on vector quantization is proposed with the purpose of protecting media distribution. The proposed JFD scheme is equipped with two encryption techniques, which are performed on the server side. The first technique encrypts plain-images using static key-trees, but the second approach uses dynamic key-trees to further simplify the first method. When the subscriber receives the encrypted images, these images are jointly decrypted and fingerprinted and are slightly different from the original images. The proposed method, to the best of our knowledge, is the first JFD method for vector quantization (VQ) compressed images. The experimental results show that the encrypted image is unintelligible and that the recovered image has desirable image quality resistant to noise interference.

► A JFD scheme for protecting image distribution with traitor tracing is proposed.
► The reason of adopting VQ is that it is particularly strong against fault tolerance.
► The first method encrypts images based on permutation and codeword substitution.
► The second uses a dynamic key-trees approach to replace permutation.