基于ASVD和BCH码的鲁棒音频盲水印算法OA

To address the issue in audio copyright protection where watermark algorithms struggle to extract or even lose the watermark when subjected to malicious desynchronization attacks such as time scaling and pitch shifting,a robust blind audio watermarking algorithm based on ameliorated singular value decomposition(ASVD)and bose-chaudhuri-hocquenghem codes(BCH codes)is proposed.At the embedding end,the watermark signal is first subjected to error-correcting coding using BCH codes.The audio,after being framed and segmented,undergoes discrete cosine transform(DCT)and ASVD de-composition.Subsequently,leveraging the insensitivity of the relative relationship between 2 adjacent segments to desynchronization attacks,the watermark information is embedded using a quantization-based singular value method.At the extraction end,the relative relationship between adjacent segments of the audio under detection is first analyzed to extract the watermark signal.Then,the original water-mark is recovered based on BCH decoding.Simulation experimental results demonstrate that the pro-posed algorithm exhibits good imperceptibility.It not only shows robustness against common signal processing attacks but also remains capable of extracting complete watermark information when subject-ed to desynchronization attacks such as time scaling and pitch shifting.