An Adaptive Robust Watermarking Algorithm for Audio Signals Using SVD

Abstract

This paper proposes an efficient watermarking algorithm which embeds watermark data adaptively in the audio signal. The algorithm embeds the watermark in the host audio signal in such a way that the degree of embedding (DOE) is adaptive in nature and is chosen in a justified manner according to the localized content of the audio. The watermark embedding regions are selectively chosen in the high energy regions of the audio signal which make the embedding process robust to synchronization attacks. Synchronization codes are added along with the watermark in the wavelet domain and hence the embedded data can be subjected to self synchronization and the synchronization code can be used as a check to combat false alarm that results from data modification due to watermark embedding. The watermark is embedded by quantization of the singular value decompositions in the wavelet domain which makes the process perceptually transparent. The experimental results suggest that the proposed algorithm maintains a good perceptual quality of the audio signal and maintains good robustness against signal processing attacks. Comparative analysis indicates that the proposed algorithm of adaptive DOE has superior performance in comparison to existing uniform DOE.