基于扩频的音频信息隐藏技术研究

发布时间：2018-03-14 07:05

本文选题：信息隐藏　切入点：音频信号　出处：《中国科学院大学(工程管理与信息技术学院)》2016年硕士论文　论文类型：学位论文

【摘要】：音频信息隐藏技术是信息隐藏技术领域的一个重要分支,在数字水印和隐蔽通信等方面有着广泛应用。基于扩展频谱的信息隐藏算法由于具有抗干扰能力强、隐蔽性好、安全性高等优点,自提出以来就一直是研究的热点。本文主要对基于扩展频谱的音频信息隐藏技术进行了深入研究。主要工作如下：1.基于统计理论对载体信号、嵌入强度与扩频序列相关检测结果之间的影响关系进行了研究,用于指导扩频序列长度和相关检测阅值的选取。2.提出了一种综合考虑同时掩蔽和后掩蔽效应,并在频域对掩蔽阈值逐点进行独立计算的方法。该方法能够充分利用听觉掩蔽特性,有利于在不失合成音频感知透明性的前提下提高隐藏信息的容量和鲁棒性。3.通过对各帧频谱进行同相修改和多帧叠加,提出了一种基于扩频调制的音频信息隐藏改进算法。算法对失同步不敏感,具有较强的鲁棒性。4.提出了一种基于正弦调制的信息隐藏算法。算法采用具有谐波关系的正弦信号对隐藏信息进行调制并将其嵌入到原始音频信号的频谱中,嵌入信息可直接采用FFT算法进行检测,计算效率高；由于有多个谐波任意可选,每个谐波均可代表一种不同的编码样式,因而隐藏容量大；另外算法在嵌入时采用感知规整,保证了合成音频的感知透明性。应用本文所提算法在不同的音频载体上进行了仿真实验,通过理论分析和仿真实验表明,本文所提算法对失同步不敏感,能够有效抵抗噪声、滤波等攻击,具有较强的鲁棒性和可靠性。
[Abstract]:Audio information hiding technology is an important branch of information hiding technology, which is widely used in digital watermarking and covert communication. The advantages of high security have been a hot topic since it was put forward. In this paper, the audio information hiding technology based on spread spectrum has been deeply studied. The main work is as follows: 1. Carrier signal based on statistical theory. The relationship between embedding strength and correlation detection results of spread spectrum sequence is studied, which is used to guide the selection of the length of spread spectrum sequence and the correlation detection reading value. 2. A comprehensive consideration of simultaneous masking and post-masking effect is proposed. The method of independent calculation of masking threshold point by point in frequency domain can make full use of auditory masking property. It can improve the capacity and robustness of hidden information without losing the transparency of synthetic audio perception. An improved algorithm for audio information hiding based on spread spectrum modulation is proposed. 4. A information hiding algorithm based on sinusoidal modulation is proposed. The algorithm uses sinusoidal signal with harmonic relation to modulate the hidden information and embed it into the spectrum of the original audio signal. The embedded information can be detected directly by FFT algorithm, which has high computational efficiency. Because there are multiple harmonics optional, each harmonic can represent a different coding style, so it has a large hidden capacity. In addition, the algorithm adopts perceptual regularity when embedded. The experimental results on different audio carriers show that the proposed algorithm is insensitive to the loss of synchronization and can effectively resist noise. Filtering and other attacks have strong robustness and reliability.
【学位授予单位】：中国科学院大学(工程管理与信息技术学院)
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP309

【相似文献】