Date Added: Sep 2011
In this paper, auditory cryptography (k, n) threshold watermarking algorithm is proposed. By binary halftone pretreatment, one-dimensional meaningful audio watermarking is pretreated to high-fidelity binary audio. The binary audio is encrypted to n cryptographic audios and embedded in scrambled host audio signal hierarchically. Less than k of the cryptographic audios give no information about the original, only synchronized playing k or more than k of the audios the original can be heard directly. Compared with the traditional audio watermarking algorithm, the proposed algorithm has high security, strong practicability and robustness in enduring common attacks.