Audio Watermarking Combined with Compressive Sampling Based on QIM and DST-QR Techniques

Abstract

Abuse is not only done to copy or distribute data but also to the digital copyright labels. There is a way to protect data by inserting or hiding a piece of certain information, namely a watermarking technique. In this paper, we propose audio watermarking with Quantization Index Modulation (QIM) method as an embedding process combined with Compressive Sampling (CS), Discrete Sine Transform (DST) and QR decomposition. Binary image is used as a watermark inserted in host audio. DST is used for transformation process from time domain to frequency domain, while QR is used to decompose onedimension matrix into two-dimension matrix. Meanwhile, CS is used to obtain the compressed watermark file which is done before the embedding process. QIM method is used to embed the watermark file to the audio host file. Simulation results indicated that the proposed audio watermarking technique has good robustness against some attacks such as Low Pass Filter (LPF), resampling and linear speed change. In addition, it provides good performance in terms of imperceptibility with Signal to Noise Ratio (SNR) > 20 dB and capacity C = 689 bps.