Method for watermarking robust audios with invariable time scale based on zero-crossing rate

Abstract

The present invention discloses a method for watermarking robust audios with invariable time scale based on zero-crossing rates, which comprises the following steps: firstly, dividing an original audio signal A into a plurality of nonoverlapping audio frames; secondly, calculating the zero-crossing rate of each audio frame, and selecting M1 audio frames with larger zero-crossing rates for embedding watermarks according to a preset threshold value T; thirdly, dividing each one of the M1 audio frames into a plurality of nonoverlapping audio segments of which the number is accordant with the total bit number of the watermarks to be embedded, quantizing the average value of nonoverlapping audio segment samples so as to embed watermark bits, and modifying sample values which correspond to each audio segment of the watermarks to be embedded, quantizing the average value of nonoverlapping audio segment samples so as to embed watermark bits; and finally, interconnecting the audio segments containing the watermarks to obtain audio frames containing the watermarks so as to form the final audios with the watermarks. The method has a simple algorithm, is applicable to all types of audio signals, and not only can effectively resist attacks of conventional signal processing, but also has strong robustness for time scale modifications in a resampling form., but also has strong robustness for time scale modifications in a resampling form.

Description

technical field [0001] The invention relates to a time-scale-invariant robust audio watermarking method based on zero-crossing rate, in particular to a robust audio watermarking method suitable for digital audio copyright protection and capable of effectively resisting re-sampling time scale modification. Background technique [0002] In recent years, with the rapid development of computer network technology and multimedia technology, quite a few audio watermarking methods have been proposed for copyright protection of audio data. However, the synchronization attack problem is still a hot and difficult problem in the field of audio watermarking. Synchronization attacks mainly include cropping and time-scale modification (Time-Scale Modification, TSM). Compared with shearing, time scale modification can stretch or shorten the length of audio signal without changing the audio quality and perception, and it is more destructive to the watermark information embedded in the audio...

AI Technical Summary

Problems solved by technology

Based on the literature "An audio watermarking method robust against time and frequency fluctuation" (R. Tachibana, S. Shimizu, Proceedings of SPIE Security and Watermarking of Multimedia Contents III, vol.4314, pp.104-115, 2001), the literature "Improving audio watermarking robustness using stretched patterns against geometric distortion" (R. Tachibana, Proceedings of IEEE International Conference on Multimedia, pp.647-654, 2002.) Improves performance against time scale modification by using multiple pseudorandom arrays to ± 8%, but its performance still depends on the tuning of some parameters

The literature "Localized audiowatermarking technique robust against time-scale modification" (W.Li, X.Y.Xue, P.Z.Lu, IEEE Transactions on Multimedia, vol.8, no.1, pp.60-69, 2006) embeds watermarks in perceptually important The local region of the method resists the pitch-invariant time scale modification attack to ±15%, but this method is only suitable for some specific music

The literature "Histogram-based audiowatermarking against time-scale modification and cropping attacks" (S.J.Xiang, J.W.Huang, IEEE Transactions on Multimedia, vol.9, no.7, pp.1357-1372, 2007) uses audio signal histograms and absolute Mean, proposes a robust audio watermarking method that is effective against temporal scale modification ±30%, however, the method is less robust to conventional signal processing attacks such as lossy compression

Images