Energy-proportion based scheme for audio watermarking

This study presents a audio watermarking that embeds information by energy-proportion scheme. By using normalised energy instead of probability, this study rewrites the entropy in information theory as an energy-proportion function. In order to guarantee the robustness of watermarks, low-frequency coefficients of discrete wavelet transform (DWT) are mapped into the domain of this function. Then, the characteristic curve of energy-proportion function (CCEP) is obtained. Moreover, some characteristics and properties of this function are analysed and proved. Based on CCEP and the properties of energy-proportion function, this study proposes a novel audio watermarking scheme. In addition, the proposed energy-proportion scheme can extract the watermark without original audio signal. Finally, performance of the proposed scheme is assessed by signal-to-noise ratio (SNR), mean opinion score, embedding capacity and BER. Experimental results demonstrate that the embedded data are robust against most signal processing and attacks, such as re-sampling, MP3 compression, low-pass filtering and amplitude scaling.