The variety of properties of different audio signals crucially affects the selection of suitable watermarking technology for satisfying the various requirements of copyright protection applications. However, comparison of audio watermarking methods has been widely ignored in the literature. In this paper, we compare the performance of three different audio watermarking methods: echo hiding, cepstrum domain, and patchwork algorithms. We investigate the no-attack situations, as well as cases when common signal processing or intentional attacks are imposed on the watermarked signal. By using modified parameters for these methods, we propose better results with respect to some previous research studies.

A comparison of some audio watermarking methods

Echo hiding is a steganographic technique that hides messages in audio.This paper proposes a scheme that improves the robustness of echo hiding.A self-synchronizing code set called the T-Codes is used to encode messagesprior to their concealment using echo hiding. This is shown to increasemessage recovery and in turn robustness of echo hiding.

Experiments with and Enhancements to Echo Hiding

In this paper, combining with the zero-crossing rate and short-time energy of the audio frame characteristics, an algorithm based on BTC, Arnold and chaos scrambling encryption for seeking the audio frame suitable for the embedded watermark is proposed. And also, the energy is dynamically adjusted to determine the embedding strength so as to self-adaptively embed watermark in the audio frame. Furthermore, the proposed algorithm, as a zero-watermarking algorithm, doesn't need to modify the original audio signal and can realize blind detection during watermark extraction. The imperceptibility of the algorithm is good, and the security is well due to the Arnold and chaos scrambling encryption. The simulation attack results show that the algorithm has good robustness.

Digital audio blind watermarking algorithm based on audio characteristic and scrambling encryption

A robust audio zero-watermark scheme which constructed the watermark in lifting wavelet domain based on chaotic modulation is presented in this paper. Since the constructed zero-watermark is not embedded into the original audio indeed, the contradiction between imperceptibility and robustness is perfectly solved. The algorithm based on coefficient maximization guarantees that the zero-watermark which constructed by low frequency coefficients represents the significant characteristics of the audio signal, experiment results show that it is robust against general audio processing such as noise addition, low-pass filtering, requantization, resampling and mp3 compression. And after the watermark is created, an initial vector is obtained by chaotic modulation, so the storage space is reduced after generating the watermark.

Robust Audio Zero-Watermark Based on LWT and Chaotic Modulation

Echo hiding is one of the prevailing techniques in audio watermarking. Currently available echo hiding technique still has some drawbacks, the detection ratio is relatively low, the robustness against many common signal processing operations of amplitude amplifying, Mp3 compression, re-sampling, filtering, time-delay and noise addition is not satisfactory. To enhance the detection ratio, we propose using the autocorrelation of power cepstrum instead of the original auto-cepstrum in that the former is more effective in the presence of attacks. Computer experiments are performed to test the validity of the proposed scheme, the results of which show the proposed detection scheme is superior to the original one especially in the presence of attacks.

A Novel Echo Detection Scheme based on Autocorrelation of Power Cepstrum

In this paper, we survey the audio watermarking. The watermarking implementation techniques are briefly summarized and analyzed. The audio watermarking is classified in three categories: patchwork in the frequency domain, echo hiding in the time domain and cepstrum domain. Experimental results show us which scheme has a good robustness against common signal processing manipulations.

M.N. Miyatake

Papers

Analysis of audio watermarking schemes