Wavelet domain self synchronization digital audio watermarking algorithm based on channel information

Abstract

With the increasing piracy and infringement of digital audio media, the copyright protection of digital audio media has attracted more and more attention. Digital watermarking is an important means to solve the problem of audio media copyright protection. The digital watermarking applied to copyright protection requires high robustness. How to improve the anti attack ability of audio digital watermarking methods, especially the robustness of audio watermarking, has become an urgent problem to be solved. A wavelet domain self synchronization digital audio watermarking algorithm based on channel information is proposed. This algorithm is a blind detection algorithm, which does not need the original audio signal. The algorithm introduces the idea of coefficient mean quantization. After the audio signal is decomposed by wavelet, the low-frequency coefficients are mean quantized and watermarked. The idea of embedding synchronization signal is introduced, and the embedding position of watermark is located by using synchronization signal to resist attacks such as translation and cropping. Experimental results show that the algorithm has strong robustness and anti-attack.