无忧论文网
当前位置: 无忧论文网 > 自然科学论文 > 计算机科学论文 > 计算机软件论文 > 基于MPEG-7视频水印的研究与实现
点击提交论文指导需求
高薪诚聘老师
基于MPEG-7视频水印的研究与实现
时间:2011-02-16 浏览次数:612次 无忧论文网
点击这里在线咨询我
计算机软件与理论
数字水印数字水印
    
    随着多媒体技术和计算机技术的发展,使得人们创作、使用多媒体数字产品变得越来越容易,互联网技术的发展极大的推动了多媒体信息的广泛传播。然而,由于数字产品复制不会引起质量下降,非法用户可以容易的获取、复制、传播数字多媒体产品,这极大的损害了版权所有者的经济利益,多媒体信息的安全问题成为一个非常重要的现实问题。
    数字水印技术(Digital Watermarking)作为数字产品版权保护的有效手段,在不影响原始数据正常使用的情况下嵌入版权信息,并且该信息不易被除去,只能由专用的水印提取算法才能检测其中的信息。因此,它成为多媒体信息安全研究的重点,其中视频数字水印更是当前研究的热点。在各种数字水印技术中,视频数字水印技术由于其自身的研究价值和潜在的经济利益而倍受各国学者的关注。然而相对于静止图像水印技术,视频水印技术的发展相对滞后。而对于MPEG-7标准下的视频水印算法更是缺乏深入的研究。
    本文提出一种针对几何攻击具有鲁棒性的数字水印算法模型,利用小波变换,将原始图像进行分解,获得原始图像的低频域,并在该低频域上进行水印的嵌入和提取的操作。与传统的傅立叶梅林变换相比,本文的算法使用图像能量中心不变点作为对数极坐标中心点,使图像经过攻击以后中心点能保持相对一致性。同时,利用傅立叶变换的可叠加性,直接对图像经过对数极坐标后产生的区域进行二维傅立叶变换,并获得变换后的幅度域,其中具体水印序列的嵌入和提取在该域完成。在图像水印研究的基础上,又提出了一种基于MPEG-7形状轮廓编码的视频水印的算法模型,该水印算法模型充分和MPEG-7的标准相结合,利用MPEG-7基于形状轮廓编码,确定水印嵌入的区域,使得数字水印的系统不再单纯的独立于MPEG-7的标准编码之外,这是针对MPEG-7与数字水印技术的有机结合做出的一个比较大胆的尝试。实验结果表明,这样的尝试利用了编码器本身的特征提取的内容,对于保证水印的鲁棒性提供了很大的帮助。
     [英文摘要]:     With the development of multimedia technology and computer technology, people can create and use the multimedia products much more easily, and communicate the digital information with each other all over the world. However, because the duplication of digital products cannot cause the quality loss, illegal users can copy and transmit the digital multimedia products, which harms the owner's economic interest greatly. The problem of the multimedia security especially the copyright problem becomes more and more important.
    As an effective measure of the digital product copyright protection, Digital Watermarking embeds the copyright information into the original data. It does not affect the universal use of the original data and the watermark information can not be erased easily, but extract it by the special extracting or examining algorithms. Therefore, it becomes the key point of the multimedia information security research. And because of its potential economic interest and the research value, the video watermarking technology attracts the scholar's attentions all over the world. But compared to the still picture watermark technology, the development of video technology appeared properly slowly.
    The objective of this thesis is to design a practical video watermarking system, and it is proposed a new image watermarking scheme which is robust to RST attacks with cropping by improving Fourier-Mellin transform based wavelet transform. This proposed scheme decomposes the original image through wavelet transform and obtains the low-frequency domain, and then embeds and detects the watermark in this domain. Unlike FMW, this proposed method uses Log-Polar map(LPM) in the spatial domain for scaling invariance, while translation invariance is provided by the use of an invariant centroid as the origin of LPM. This invariant centroid makes sure that the origin of LPM after geometrical attack is as same as before. For this, its calculation method is proposed. Also according to the DFT property of accumulating, 2-D DFT is performed on the LPM image and the magnitude spectrum extracted to provide a domain that is rotation invariant. The resulting domain, which is invariant to RST, is then used as the watermark-embedding domain. Experimental results demonstrate that the proposed scheme is robust to RST attacks.
    Based on the robust watermarking scheme, it is taken practice into video watermarking model system. Taking advantage of the shape contour descriptor in MPEG-7, it can ensure the region where the watermark will be embedded in this video watermarking model system, which combines the watermarking model with MPEG-7 standard. And this is the properly audacious attempt to these two aspects. Experimental results demonstrate that this video watermarking model is feasible in practical applications.    
关于我们 | 老师招聘 | 版权声明 | 联系我们 | 付款方式 | 返回顶部 | 

COPYRIGHT ©2001 - 2013 51LUNWEN.NET. ALL RIGHTS RESERVED.
【免责声明】:本网站所提供的信息资源如有侵权、违规,请及时告知
无忧论文网提供毕业论文指导 硕士论文指导服务