一种健壮的数字音频水印时域算法

A Robust Digital Audio Watermarking Time-domain Algorithm

提出了一种健壮的数字音频水印时域算法。把音频信号划分为若干个包含相同采样点的段,每一段划分为若干个包含相同采样点的节,对每段前二节的能量进行比较,结合水印信号及HAS(人类听觉系统,Human Auditory System)的掩蔽特性,采取不改变或缩小音频信号能量的方法,在数字音频中嵌入水印。实验证明,该算法听觉相似性好,健壮性强,可盲水印检测。

This paper presents a robust digital audio watermarking time-domain algorithm.The audio signal has been divided into several passages which including the same sample,and the passage divided into several sections which including the same sample.Energy of the first two sections of each passage has been compared,according to watermark bit and characteristic of HAS(Human auditory system),and adopting the way of not changing or reducing energy,watermark has been embedded into the audio signal.The experiment result shows that the algorithm has the better inaudibility and more robustness and detect is blind watermark.