基于能量选择的DCT域音频数字水印算法

An Audio Digital Watermark Algorithm in DCT Domain Based on Energy Selection

文中提出了基于能量选择的离散余弦变化域的音频水印算法。该算法根据人类听觉系统掩蔽特性和离散余弦变换能量不变的特性,利用在音频信号能量较大处插入水印对音频信号影响较小的特点,将二维的图像降维成一维的二值序列,通过音频信号时域序列的能量选择在离散余弦变换域低频系数中自适应地寻找最佳的水印图像插入点。实验表明,文中提出的算法不仅有较好的隐蔽性,而且对于数字信号处理操作中的加性高斯白噪声、滤波等具有较好的鲁棒性。

An audio digital watermark algorithm in DCT domain based on the energy selection is proposed in this paper. A right position in the low frequency coefficients of the DCT domain is adaptively found by the energy selection of the time-domain sequence based on masking properties of the human auditory system （HAS） and the energy invariant characteristic of discrete cosine transform （DCT）. U- sing the feature of inserting an audio digital watermark at the larger energy place of the audio signal will cause the less impact on the audi- o signal and converting the two dimensional image to a one-dimensional binary sequence. Experiment results show that the proposed al- gorithm is not only imperceptible, but also robust to common operations to digital audio signal, such as additive white Gaussian noise, fil- ter and so on.