基于小波变换的语音隐藏方法研究-通信与信息系统专业论文.docxVIP

基于小波变换的语音隐藏方法研究摘要 基于小波变换的语音隐藏方法研究 摘要 基于小波变换的信息隐藏技术是近年来研究的热点。本文提出了两 种基于小波变换的语音隐藏方法，把一段语音作为秘密信息嵌入一段数 字音频信号中，其目的是用音频信号作伪装来传递秘密的语音信息或把 语音信息作为一种标记来保护音频作品。 本文介绍的第一种方法根据人类听觉系统的特点，利用小波对语音 信号进行压缩，增加了数据的嵌入量。载体音频信号小波变换后，选择 部分小波系数用适当的嵌入参数嵌入信息。实验表明该算法的隐蔽性好， 隐藏的数据量大，对常见的信号处理操作具有较强的稳健性，符合掩蔽 通信的要求：第二种方法在信息提取时不需要原始载体信号。载体信号小 波变换系数的偶数点用相邻系数的平均值替换后用来嵌入信息。替换的 系数经过嵌入数据修改后更接近原始的系数。实验表明该算法在隐藏的 信息量不大时对常规的数据操作譬如压缩、滤波、重采样，加白噪声等 具有较强的稳健性，适合把秘密语音作为标记嵌入音频信号中达到版权 保护等目的。 关键词：信息隐藏数字水印掩蔽通信小波变换 作 者：彭静玉 指导老师：赵鹤鸣 基于小波变换的语音隐藏方法研究 基于小波变换的语音隐藏方法研究 Abstract Research of Speech Hiding Scheme Based on黝矿elet Transform Abstract Information hiding techniques based on wavelet transform is a hot research field in last few years．This paper proposes two kinds of speech hiding algorithm based on wavelet transform．It embeds a speech as the secret message into a digital audio signal in order to transmit secret message under the camouflage of the audio signal or protect the audio product by using the speech signature as a mark． The first method described in this paper makes use of the human auditory system(HAS)characteristics and the wavelet to compress the speech signature，SO it increases the capacity．A proper parameter and some wavelet coefficients of the cover audio signal are selected to embed information．The experimental results show that the algorithm can hide high capacity of date imperceptibly and it is robust to routine operations on digital signal．The second method doesn’t need the original cover signal when extracts the secret dates．The even wavelet COCfficients of the cover signal are substituted by the mean value of the aajacent coefficients and used to embed information．After being modified by embedded dates the substituted coefficients are closer to original coe艏cients．The experimental results show：when the capacity is not SO large，the algorithm is robust to many operations to digital date，such as compression，filter,re—sampling，white noise etc．It is adapt to embed secret speech as a mark into audio sig