Audio watermarking based on inverse difference decomposition

A new approach for audio watermarking in the phase-frequency domain is presented, based on the inverse difference decomposition (IDD) with complex Hadamard transform (CHT). The main advantages of the method, compared with others based on the discrete Fourier and wavelet transforms, are the absence of quantization noise, the lower computational complexity and the ability for the owner of the audio contents and authorized distributors to insert different watermarks in the protected audio signal. Additionally, there is watermark transparency, high resistance against fraud and "pirate" attacks, and the capability for exact watermark extraction without using the original audio signal.