Digital Audio Watermarking Using Psychoacoustic Model and CDMA Modulation

Digital watermarking is used to insert information (a signature) in a computer document. The addition of the signature must be imperceptible and undetectable by any system ignoring its mode of insertion. In particular, it must be completely invisible to the human eye. This method is different from cryptography, which hides a message by making it unreadable. Digital watermarking now includes other data within music signals. Audio watermarking consists in embedding inaudible information in an audio signal. The watermarking system must guarantee one transmission of the inaudible, reliable and robust information face a set of disruptions. In this goal, we propose a new strategy of insertion adapted to a watermarking system. This strategy permits to construct an inaudible watermarking and of maximal hardiness to the addition of a noise.