New secure and robust audio watermarking algorithm based on QR factorization in wavelet domain

Abstract

Digital watermarking consists in embedding imperceptible information into a host signal. It has been proposed to solve problems as varied as the protection of the copyright, content authentication, fingerprinting and broadcast monitoring. This paper presents a new approach for audio watermarking using the QR factorization in wavelet domain. This approach is based on embedding a watermark binary image in the R matrices of low frequency blocks DWT coefficients of audio signal. In this algorithm, the watermark is embedded by applying a Quantization Index Modulation (QIM) process on the determined optimal sample for each matrix R. The watermark can be blindly extracted without the knowledge of the original audio signal. Experimental results show that the proposed audio watermarking scheme maintains high quality of the audio signal. Signal to Noise Ratio (SNR), Log Spectral Distortion (LSD) and Mean Opinion Score (MOS) are about 40 dB, 0.37 dB and 4.84, respectively. Moreover, the scheme is quite robust against common signal processing attacks such as noise addition, filtering and MP3 compression. In addition, this method ensures a secure extraction process by using a private key, making it suitable for secure applications such as copyright protection.