【作者】 占徐光；

【导师】 罗斌；

【作者基本信息】 安徽大学 ， 计算机应用技术， 2010， 硕士

【摘要】 矢量量化是一种高效的数据压缩技术,具有压缩比大且解码简单因此广泛应用于图像压缩、语音编码等多个领域。数字水印技术是将序列号、文字、二值图像等信息嵌入到多媒体数据中,目的是为了多媒体信息的版权跟踪和版权保护,特别是在图像、视频中得到了广泛的应用,成为多媒体信息安全研究领域的一个热点问题。在矢量量化基础上研究数字水印,就是在压缩域上研究数字水印,基于矢量量化的数字水印技术基本思想是利用匹配输入矢量集的码字索引表传输版权保护的信息。伴随着数码产品的更新换代,高清数码相机进入日常生活中,.彩色图像成为重要的数字水印研究对象,同时随着3G时代的到来,智能手机的普及与主流智能手机的相关增值业务的出现,基于智能手机的矢量量化数字水印技术必将成为一个研究的重点。本文主要是针对基于矢量量化的数字水印技术的研究,分析了矢量量化关键技术中的码书设计与码字搜索。选择彩色图像为研究对象,提出了一种在彩色图像YUI色彩空间的基于矢量量化信息压缩的DCT域的数字水印算法。依据智能手机平台的特点选择一种特殊码书设计与码字搜索的策略,同时将基于图像检索的理论融合到该系统中,分别设计了最优码书生成系统与自识别VQ系统。主要工作内容及具体改进点如下：(1)简单介绍Symbian智能手机操作系统,重点研究了码书设计与码书搜索算法。在码书设计算法中分析了传统的LBG算法与PNN算法的优劣性,在后面的码书设计中将LBG与PNN算法结合；在搜索算法中比较了FS与EENNS算法与IEENNS算法优劣性,从而选择IEENNS作为智能手机码书搜索算法(2)选取彩色图像为研究对象,分析了常见的几种彩色图像的颜色模式以及相互之间的转换,最后提出了一种在彩色图像YIQ色彩空间的基于矢量量化信息压缩的DCT域的数字水印算法。增大了水印的信息容量与水印的安全性,不同层次的水印能够分别抗JPEG与VQ的压缩,达到了嵌入多重水印,适应于多用途。(3)在矢量量化数字水印技术的实际应用中,选择Symbian Os作为智能机的操作系统,以Nokia智能机为研究对象。本章分为两个子系统,第一个是最优码书生成系统,第二个是自识别VQ系统,前一个系统为后一个系统的基础,本文分别给出了两个系统的设计原型更多还原

【Abstract】 Vector quantization is a highly efficient data compression techniques,as the compression ratio is large and simple it was widely used in decoding the image compression, voice coding and other fields. Digital watermarking technology which embedded the serial number, text, binary images and other information into the multimedia data, multimedia information is aimed at tracking and copyright protection of copyright, especially in the image, the video has been widely used as multi-media information security area of research is a key point. The study of digital watermarking which base on vector quantization, that is, to study in the compressed domain digital watermarking. Vector quantization of digital watermarking technology basic idea is to match the input vector set to use the code word index table transfer of copyright protected information. With the upgrading of digital products, high-definition digital camera come into the daily life, color image digital watermarking has become an important object of research, at the same time as the 3G era, the popularity of smart phones, the mainstream smart phones associated with emergence of value-added services, based on Smartphone vector quantization digital watermarking technology will become a focus of the studyThis article is based on vector quantization for digital watermarking technology, research, analysis of the key technologies in the vector quantization codebook design and code word search. Select color images as research object, presents a color image YIQ color space information based on vector quantization of the DCT compressed domain digital watermarking algorithm. Based on the smart phone platform’s characteristics to select a particular codeword and codebook design and search strategy, while the theory-based image retrieval integrated into the system were designed to generate the optimal codebook VQ system with self-recognition system. Major work to improve the content and the specific points are as follows: (1) Briefly introduce Symbian smart phone operating system, focusing on the codebook design and codebook search algorithm. With the analysis advantage and weakness of the traditional LBG algorithm and PNN algorithm, combined LBG codebook design and the PNN algorithm as the selected algorithm to generate codebook; In the search algorithm to compare advantage and weakness of the FS,with the EENNS algorithm and IEENNS algorithm, In the end Select IEENNS as smart phones codebook search algorithm(2) Select a color image as the object of study, analysis of several common color model and color image conversion.Thereby presents a color image YUI color space information based on vector quantization of the DCT compressed domain digital watermarking algorithm. Increase the information capacity of the watermark and watermark security, different levels of the watermark can be resistant, respectively JPEG compression with VQ reached the embedding of multiple watermarks, adapted to multi-purpose.(3) In the vector quantization of the practical application of digital watermarking technology, select Symbian Os operating system as an intelligent machine to Nokia smart phone for the study. This chapter is divided into two subsystems, the first one is the optimal codebook generation system, the second is the self-recognition VQ system, the previous system after a systematic basis, this paper gives the two systems were prototypes更多还原