【作者】 刘锋；

【导师】 沈永增；

【作者基本信息】 浙江工业大学 ， 控制理论与控制工程， 2004， 硕士

【摘要】 伴随着网络技术和多媒体技术的飞速发展，多媒体数据逐渐成为人们获取信息的重要来源，并成为人们生活的重要组成部分。因而，如何保护多媒体信息的安全成为国际上研究的热门课题。本论文主要探讨有关数字图象信息安全中的数字水印技术。数字水印技术作为版权保护的重要手段，得到广泛的研究和应用。它通过在原始数据中嵌入秘密信息—水印(watermark)来证实该数据的所有权或完整性，以此来抵制对数字作品的盗版或篡改等。 本人在导师的指导下，经过努力，在数字水印技术方面得到了以下的结果： 1．提出了一种基于DCT和SVD的一种灰度水印嵌入算法：现在大部分的数字水印技术所嵌入的水印一般是随机信号或者是二值图象，那么用有意义的灰度图象作为水印信号的嵌入技术很少见，本人所提出的这种算法对这方面的技术进行了尝试，提出了一种基于DCT和SVD的灰度水印嵌入算法，这种算法的先进之处就是嵌入的水印信息是信息量很大的灰度图象。经过实验仿真，把该算法能够经受住噪声、中值滤波、压缩、高通滤波、低通滤波等的处理，具有很强的鲁棒性。 2．提出了一种基于DWT和DCT的盲的数字水印嵌入技术：现在大部分的数字水印技术在检测水印时都需要原图象，但是这样的水印嵌入技术实用性不强，为了增加水印嵌入技术的实用性，本人提出了一种基于DWT和DCT的盲的数字水印嵌入技术，这种技术在检测水印时不需要原图象，而且水印信号是一种有意义的二值图象信号。该算法首先对图象进行小波分解，对LL图象按照一定大小进行分块，然后对每一块进行离散余弦变换，把水印图象的值嵌入到变换后的系数上，最后再进行逆变换。把该算法能够经受住噪声、中值滤波、压缩、高通滤波、低通滤波等的处理，具有很强的鲁棒性更多还原

【Abstract】 The digital media has become a main way for information communication along with the wide use of network. At the same time, multimedia information security problems including copyrights piracy emerged as side effects of the popularity of digital representation and distribution over network. Information hiding, also called Sicganography , digital watermarking technology provide novel ways to solve these security problems for multimedia information. In this dissertation, we study the technologies and application of multimedia information security in the area of digital watermarking. The main contributions of this dissertation are:A Newly Watermarking Algorithm With Gray-level Image based on DWT And SVD is proposed: Now, most of the watermarking algorithms hide the random signal or two-value image into the image; only a few watermarking algorithms hide the gray-level image into the image. In my dissertation, we study the algorithm that hides the gray-level image into the image and gain some result. The algorithm that we study is based on DWT and SVD. The main merit of the algorithm is the capacity of the watermarking information that is hid into the image is much more than the capacity of the random signal and two-value image. The experiment based on this algorithm demonstrate that the watermarking is robust to the common signal processing techniques including compressing、 noise、low pass filter、high pass filter and median filter.A blind watermarking for digital image based on DWT and DCT is proposed: Now , in the most of the watermarking algorithms , the detection of the watermarking need the original image, so the practicability of the watermarking is not good . In order to strengthen the practicability of the watermarking,we proposed a blind watermarking algorithm for Digital Image Based On DWT and DCT. This algorithm manipulate the original image by DWT, then adds watermark onto the important wavelet coefficients located in the low frequency sub-band which is splitted into blocks and manipulated by DCT. The experiment based on this algorithm demonstrate that the watermarking is robust to the common signal processing techniques including compressing、 noise、low pass filter, high pass filter and median filter.更多还原