Audio Watermarking Based on Empirical Mode Decomposition

Abstract

an audio watermarking based on Empirical mode decomposition is proposed. The watermark is embedded in the final residual because of the intrinsic feature of the final residual. The audio signal is divided into set of frames and by using the empirical mode decomposition (EMD), frames are decomposed into set of intrinsic mode functions (IMF) and a final residual. The watermarked data is embedded in the final residual and the algorithm works by shifting each element of the final residual for make its sum greater or less than 0. This algorithm does not change the property of the final residual after embedding the watermark bit on to the final residual because of the watermark is robust against various types of attacks.