Blind audio watermarking algorithm based on DCT, linear regression and standard deviation

• Published in: Multimedia tools and applications Vol. 76; no. 3; pp. 3343 - 3359
• Main Authors: Jeyhoon, Mahdi, Asgari, Mohammad, Ehsan, Lili, Jalilzadeh, Seyedeh Zahra
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.02.2017; Springer Nature B.V
• Subjects: Algorithms; Blinds; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Digital broadcasting; Digital rights management; Digital signatures; Digital watermarks; Discrete cosine transform; Fourier transforms; Frames; Gaussian process; Intellectual property; Methods; Multimedia Information Systems; Regression; Regression analysis; Signal processing; Special Purpose and Application-Based Systems; Spread spectrum; Standard deviation; Studies; Watermarking; Wavelet transforms; United States > US; Audio watermarking; Blind; Standard deviation; DCT coefficients; Linear regression
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-016-3934-9

This paper proposes a blind audio watermarking algorithm to embed data and extract them by changing the Discrete Cosine Transform (DCT) coefficients. The key idea is to divide the selected frequency band of DCT into short frames and change the samples of each frame based on the watermark bits that are embedded in. The proposed idea uses linear regression and standard deviation to extract watermark bits. The experimental results show that the method has a high capacity about 3000 bps data payload, without significant perceptual distortion. Moreover, this idea provides robustness against common signal processing attacks such as Additive White Gaussian Noise, Resampling, Re-quantizing and Echo.