Research on DFT Domain Digital Audio Watermarking Algorithm Based on Quantization

A novel audio watermarking algorithm based on neighboring samples

“…Watermarking techniques [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15] proposed during the last ten to twelve years can be divided into two broad categories, namely the time-domain based group ( [9] [12] [15]) and the transform-domain based group. The transform-domain based works ([1]- [8], [10]- [11], [13] - [14]) mainly use cepstrum domain transform, discrete Fourier or frequency transform (DFT), discrete cosine transform (DCT), discrete wavelet transform (DWT), etc.…”

“…The transform-domain based works ([1]- [8], [10]- [11], [13] - [14]) mainly use cepstrum domain transform, discrete Fourier or frequency transform (DFT), discrete cosine transform (DCT), discrete wavelet transform (DWT), etc. [1] inserts watermarks into the cepstral components to hide a narrow band signal in a wideband channel.…”

“…The QIM based techniques [3], [4], [5], and [11] all suffer from intelligent determination of the quantization steps. Other embedding algorithms also suffer from defining the amount of distortion when watermarks are embedded.…”

A Robust Audio Watermarking in Cepstrum Domain Composed of Sample's Relation Dependent Embedding and Computationally Simple Extraction Phase

“…Both algorithms proposed in [30,31], need to set optimum value of quantization parameter to make the scheme robust and imperceptible. Another computationally less complex DFT domain-based audio watermarking scheme using quantization was proposed by Tan W. et al [32]. Though, they reduced the mean square error to minimize embedding distortion, they did not provide any criteria to determine the optimal quantization step.…”

“…Research interests for watermarking in audio are new compared to that of video [2] and image [7]. The embedding schemes for audio watermarking can be classified either based on domain transformation (time domain-based scheme [8][9][10][11][12] and transform domain-based schemes [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32]) or based on the necessity of host signal during extraction (blind schemes and non-blind schemes).…”