基于音频数字水印技术算法的研究

【作者】 梁娟；

【导师】 耿国华；

【作者基本信息】 西北大学 ， 计算机软件与理论， 2004， 硕士

【摘要】 随着网络技术和多媒体技术的发展，数字水印技术成为数字版权保护的行之有效的办法。音频水印技术是应用于音频领域的水印技术，该技术通过向音频文件中嵌入秘密信息已达到版权保护、掩蔽通信的目的。本文就音频文件的版权问题，展开了音频数字水印的研究。 本文首先实现了时域水印算法。通过将时域幅值进行排序，把水印信息嵌入在幅值较大的位置。经过实验分析，时域水印算法鲁棒性较差，但其算法简单，嵌入数据量大，因此仍然使用广泛。 为了提高水印鲁棒性，本文又对频域水印算法进行了研究和分析。在基于DCT的数字水印算法中通过寻找特征区域将水印嵌入在能量最大的离散余弦系数中；在基于DVVT的数字水印算法中，利用db1小波分解取得音频的中频分量，通过乘法准则进行水印信息的嵌入。 最后，经过对已有算法的分析，提出了利用心理声学模型1自适应调节量化步长的基于DCT和DWT相结合的数字水印算法，提高了水印的不可感知性和实用性。更多还原

【Abstract】 With the development of digital multimedia technology, digital watermark has become a popular topic in the field of digital copyright protection..Audio digital watermarking technique is application on the audio field,with such technique we can embed secred information into audio files,so as to get’copyright protection’and’cover communication’etc...This paper focus on the audio digital watermarking technique in order to protect audio file’s copyrightFirst,I completed the temporal domain digital watermarking algorithm.The information of watermarking were embedded in the largest value of the amplitude by sorted. This algorithm was proved that is weaker in robust,but it is sa mple and embedded large data.In order to improve the robust of watermarking algorithm, this paper analyzed the frequency domain watermark algorithm. Embedded the watermark information into intermediate frequency in the DCT-based watermark algorithm by seek the feature field. This paper chose the dbl wavelet transform to get intermediate frequency in the DWT-based watermarking algorithm by multiplication regulation.Furthermore,this paper put forward a automatic and adaptive algorithm that can calculate the suitable threshold. This suggested algorithm makes use of the frequence analysis of the psychoacoustic model to make the embedded watermark inaudible.This idea is used in watermark embedded algorithms based on DWT and DCT combinition.更多还原