SVD based Audio Watermarking Algorithm Using Dual Watermark for Copyright Protection

Flexibility of Internet technology gives rise to concern about the protection of digital data. Digital audio watermarking is the art of hiding important data in Digital Audio. This research paper deals with a new methodology which helps to enhance the robustness of Digital Audio Watermarking technique based on Discrete Wavelet Transform (DWT) and singular value decomposition (SVD).

Objectives: To develop an audio watermarking algorithm in transform domain and improve the performance of proposed algorithm against common signal processing attacks such as filtering, adding noise, resampling, re-quantization, & volume scaling, etc. using subjective & objective measures.

Findings: In this approach, multiple watermarks (Image -64 X 64 pixels and text) are embedded into the approximate coefficients of Discrete Wavelet Transform of an audio signal. Haar wavelet is used for decomposition while SVD is applied to get the Eigen values of approximate coefficients of audio signal. Experimental work demonstrates that watermarked audio has same perceptual quality as that of original one. The performance of the proposed work is evaluated with the help of objective measures such as Signal to Noise ratio, Peak signal to Noise ratio, Bit error rate, Normalized Cross correlation, etc. Additionally, simulation results on audio signals prove that the capacity of the proposed scheme is considerably better in comparison with the existing watermarking methods. Furthermore, the proposed work is applied for different audio signals such as classical, instrumental, flute, pop & rock to check the robustness against various attacks.