SNR Based Audio Watermarking Scheme for Blind Detection

Abstract

Audio watermarking method presented in this paper embeds the watermark data in the wavelet domain and performs the blind detection of the watermark. To embed the watermark the additive watermark embedding approach is used in which scaling parameter is computed based on the SNR (signal to noise ratio). The technique is implemented in DWT domain. 32 times 32 binary image is used as the watermark data. The SNR based adaptive watermarking scheme[6] and the proposed scheme for blind detection are implemented their results are presented here. The techniques are tested against different signal manipulation processing such as filtering, resampling, compression, cropping and requantization. The host audio signal used foe the experimentation is the Indian musical signal. It is observed that the technique is resilient against various signal manipulation attacks.