【作者】 刘海涛；

【导师】 杨义先；

【作者基本信息】 北京邮电大学 ， 信号与信息处理， 2008， 硕士

【摘要】 伴随着网络技术和多媒体技术的飞速发展,多媒体数据已经成为人们获取信息的重要来源,如何对多媒体信息进行数字版权保护成为国际上研究的热门课题。数字水印技术作为版权保护的重要手段,得到了广泛的研究和应用。它通过在原始数据中嵌入秘密信息—水印(watermark)来证实该数据的有效性或完整性,以此来抵制对数字作品的盗版或篡改等。目前在数字音频水印的研究中,已提出了一些算法,但是随着多媒体技术的发展,水印技术也期待着更大的突破。本文首先阐述了本课题的研究背景和意义,并介绍了数字水印问题的提出及相关概念、原理与分类。接着,专门针对音频文件,介绍了数字音频水印算法的基本原理、实现的框架、常见的攻击与对策。第三章,结合声音的特性以及人耳的听觉特性,在介绍常见音频文件格式的基础上,区分为时域、变换域、压缩域,从不同的领域分别介绍现有的数字音频水印技术特点,并进行一定的分析与评价。接下来的两章,在介绍MID、WAV音频格式的基础上,深入的研究和分别探讨了水印数据的嵌入、水印数据的提取与验证各个算法环节。算法设计中,结合不同的应用目的与格式特点,通过实验验证其正确性。同时,算法还具有提取水印数据信息时独立盲提取的优点。每种水印算法后,都对其进行了鲁棒性、隐蔽性、数据容量方面的测试。实验结果表明在隐蔽性方面,两种格式的算法嵌入水印数据而引入的噪声完全位于人耳的听觉掩蔽曲线值以下,不会影响人耳的主观听觉。在数据容量方面,不同的音频格式下,音频信号嵌入的水印数据量可以达到相当大的一个容量。在鲁棒性测试上,本文的算法针对不同的音频格式,来验证具有鲁棒性。检查的同时,又提出了本算法相应的优点与不足,为今后更好的设计提出了一个思路。文章的最后对本课题今后的改进方向作了阐述,并介绍了数字水印技术的发展趋势。更多还原

【Abstract】 With the rapid technology development in network and multimedia, multimedia data has become an important source to people to access information. It has become an international study of the topical issues on how to protect the digital multimedia information copyright. Digital watermarking technologies are an important means of copyright protection, and have been widely studied and used. In order to resist the digital works’ piracy or tampering, the multimedia secret data -Watermark are embedded into the original information to confirm the validity or integrity of the multimedia file. At present in the Digital Audio Watermarking fields, a number of algorithms have been proposed, but with the development of multimedia technology, the watermark technology is also looking forward to even greater breakthroughs.First, this paper introduces the background and it’s Significance, and gives a brief introduction on the purpose of digital watermarking and its related concepts, principles and Classification. Then, the fundament and general framework of watermark are stated and basic attacks and countermeasures are listed. In Chapter III, with the voices of people and of the hearing ability, divided by the time Domain, transform domain, compressed domain, digital audio watermarking technology features from different fields are introduced on the introduction of common audio formats, and some analysis and evaluation are given. In the following two chapters, on the basis of introducing WAV, MID audio formats, In-depth research of the watermark data’s embedding, extraction, and verification are studied. With different application purposes and format features, embedded arithmetic is designed. Through experiments, watermark algorithm’s accuracy is validated. Meanwhile, the Watermark is extracted without the knowledge of original audio signal.Furthermore, in experiment, the paper evaluates perceptibility, robustness and the information capacity of the watermark algorithm. From the result, according to some subjective tests, the watermarked audio signal has good audio quality comparable to the original one. The noise energy from watermark embedding is totally under human masking threshold. In data capacity, among different audio formats, the watermark data embedded into audio signal can be achieved quite a large quantity. In the robustness test, the algorithms for different audio formats are run to verify robust. At the same time as inspection, for a better improvement of the designate, advantages and disadvantages of the algorithm are discussed.Finally, the aspects need to be improved and the develop trends of digital watermarks are discussed.更多还原