Robust audio watermarking scheme based on fractional Charlier moment transform and dual tree complex wavelet transform

The popular embedded watermarking method in transform domains such as DWT, LWT, DCT, and SVD achieves copyright protection of audio signals by embedding watermarks into the original audio signals. The main weaknesses of the existing methods include low payload and low robustness against common audio processing operations and attacks, especially, vulnerability to shifting attack. This paper proposes a new robust, secure, and blind audio watermarking scheme based on a novel hybrid transform which is a combination of the dual tree complex wavelet transform (DTCWT) and the fractional Charlier moment transform (FrCMT). After segmenting the original audio signal into several segments of the same length, the 3-level DTCWT is applied on each segment using filters designed specifically to have shift invariance, then FrCMT is applied on the approximation coefficients of DTCWT. Finally, a sequence of watermark bits is embedded in FrCMT coefficients by quantizing their relative energy. The fractional order in FrCMT is chosen in our scheme in a way to obtain high imperceptibility and robustness against most common audio processing operations and attacks. Moreover, it is used as a security key to ensure a high security level of the proposed scheme. The comparative analysis confirms the high robustness, security, and efficiency of the proposed watermarking method.