Research On The Invulnerability Of Audio Digital Watermarking

In the context of the mobile Internet era,audio data sharing brings convenience to people,and it is also easy to cause digital copyright tampering and misappropriation,so digital watermark technology should be born.The study of the invulnerability of audio digital watermarks is critical to the security of data.Through research and analysis of existing literatures,it is found that the hotspots of audio digital watermark research mainly include: in the embedding of watermarks,the selection of watermark intensity and position,there is no unified quantization standard and mathematical model;in addition,the detection of watermark is also The hot topic of current research;digital watermarking technology is widely used in image,but it is rarely used in audio field.This paper studies the security of audio digital watermarking,especially the research on its invulnerability.The work done in this paper mainly includes the following parts:1.The intensity of the watermark in the audio signal is selected.The guiding principle is based on the imperceptibility of the human ear;the threshold selection of the existence of the watermark is experimentally analyzed,and the threshold value cannot be considered too much,otherwise the watermark will be affected.If the threshold is too small,the actual watermark-free signal will be detected by mistake.2.For the problem of invulnerability,the principle of spread spectrum communication and the classical spread spectrum algorithm model are studied.Because the spread spectrum algorithm has the advantages of good robustness and good encryption performance,the watermark is encrypted by M sequence,and the addition is used.The method of watermarking and despreading is used to detect the watermark;the experiment verifies the invulnerability of the spread spectrum algorithm.The attack method uses resampling,cropping and adding noise to the watermarked audio.Experiments show that the algorithm is adding The invulnerability of both noise and shear is effective.3.In order to further improve the invulnerability of audio digital watermarking,an audio digital watermarking model based on wavelet transform algorithm is proposed.The model is verified by resampling,shearing and noise-adding the watermarked audio.The anti-destructive performance,experiments show that the algorithm is effective in both the noise and shear resistance.