Research On Audio Digital Watermarking Based On Energy Difference Of Wavelet Coefficients

As an effective digital copyright protection and data security technology, digital watermarking is a foreland subject in the field of information security. At present, the original digital works are changed in the process of embedding watermarks, which inevitably brings the conflict between the robustness and imperceptibility of digital watermarking. Therefore, the embedding process that does not modify any data of original works is expected, in order to enhance the robustness and transparence.On the basis of analyzing the principles and the advantages and disadvantages of several basic methods on audio digital watermarking, two algorithms based on discrete wavelet transform are proposed, with two-dimension binary image as watermark information and audio data as imbedded target.In order to improve the robustness of watermarking, one method based on energy difference of wavelet coefficients is presented. The audio signal is enframed and decomposed by wavelet transform. The coefficients of approximation and those of detail are modified to embed secrete information. Watermarks can be retrieved without the help of original audio signal. Experiment results prove that this algorithm is robust to low-pass filtering, re-quantization, echo jamming and so on.For guarding against the damage to audio signal during embedding, an audio zero-digital watermarking algorithm based on energy difference of wavelet coefficients is proposed. Original audio signal is decomposed by wavelet transform and the energy differences of the approximation coefficients and detail's are collected and classified into two classes according to the relationship between the energy differences and the thresholds. In this way, the watermarks are embedded and extracted. The restoration of the linear scaling is used to resist the influence of the linear scaling on time axis. The experimental results show that this algorithm is robust to many operations and attacks on digital audio signal, such as low-pass filtering, re-sample, re-quantization, echo jamming and MPEG encoding.