节点文献
数字水印在有线数字电视监测系统中的应用研究
【作者】 黄永弟；
【导师】 张重雄；
【作者基本信息】 南京理工大学 ， 通信与信息系统， 2010， 硕士
【摘要】 有线数字电视是一个从节目制作、信号传输到信号接收处理显示完全数字化的系统,是融计算机、通信和电子技术于一身。与传统的模拟电视相比,数字电视具有许多无可比拟的优点。然而,有线数字电视的监测系统仍然沿袭着模拟电视的管理方式,在技术手段上还没有新的突破。应对一些对有线数字电视系统的攻击和破坏,如非法信号的插入等攻击方式,仍然以人工监测干预为主,缺少自动化的手段。数字水印技术是当今一种新型实用的信息隐藏技术,近年来得到了极大的发展。将数字水印技术应用到多媒体信息中来实施保护,是当前国内外一种行之有效的办法。视频水印是数字水印技术研究运用中最多的一项技术。本文在分析了国内外数字水印技术及其相关文献资料的基础上,将视频水印技术应用到有线数字电视中来,提出了一种将视频水印技术应用到有线数字电视监测系统方案,对针对有线数字电视的非法信号插播等攻击方式进行跟踪检测,确保有线数字电视的安全播出。在监测系统方案中,本文采用一种新的视频水印算法,即在压缩视频域内基于宏块变长码域VLC行程码参数替换的算法,该算法不需要对视频流完全解码与再解码过程,从而避免了一些计算复杂性较高的操作,如DCT变换和运动补偿等,满足了视频水印的实时性等要求。为了实现视频水印的盲检测,本文对视频水印算法进行了一些改进,采用了扩展m序列作为水印进行,主要是基于扩展m序列有着良好的均衡性,便于产生好的参考点。通过实验结果表明,该算法对视频水印的各种要求均能很好地满足,对监测方案的实施具有一定的实用价值。
【Abstract】 DVB-C(Digital Video Broadcasting-Cable) system is a production from programming, signal transmission and processing signals to display completely digital TV system and it consist of computer, communication and electronic technology. Compared with conventional analog television, DVB-C has many advantages which haven’t. But the monitoring system of DVB-C is still same with the analog television. It lacks automation way to deal with illegal attack to itself, such as illegal signal insertion attack.Digital watermarking is a new practical information hiding technology, which develops great fast recently. It is an effective way to using digital watermarking to protection multimedia information at home an abroad. Video watermarking is one of digital watermarking technology, which used most widely. This paper analyzes digital watermarking technology and the relevant literature of domestic and international, propose a solution which use video watermarking technology to DVB-C monitoring system in order to solve the problem of automation monitoring.In the monitoring system solution, we proposed a new video watermarking algorithm which is based on replacing the parameter of macro block variable length codeword run field in the compressed video domain. This algorithm doesn’t require decoding or encoding of the video stream completely, thus avoiding the high complicated computation of some operations, such as DCT transform and motion compensation, to meet the real-time requirement of video watermarking. In order to achieve video watermarking blind detection, the video watermarking algorithm uses the extended m-sequences as embedded watermark because of the balance property of extended m-sequences adequately to generate a stable reference point. The experiment results show that the algorithm satisfies ht various performance indicators of monitoring system.