以資料隱藏為基礎之視訊版權保護與強健性傳輸

Data Hiding-based Video Copyright Protection and Error Resilient Transmission

陳建儒(Jian-Ru Chen)

指導教授 ： 呂俊賢 ； 范國清

國立中央大學/資訊電機學院/資訊工程學系/博士(2006年)

摘要

近年來，由於資料數位化的發展以及網路的流行，資料可以用數位的方式儲存和傳輸，數位多媒體比以往類比式多媒體的優點是容易儲存、複製，而且複製品與原品一模一樣。透過網路，人們對於資訊的取得變得方便又迅速，傳遞也變得更容易。雖然人們可藉著網路下載數位資訊來使用，然而數位化的資訊極易被竄改與複製，這其中便牽涉到智慧財產權的侵權問題，因此數位資訊保護的相關問題便因應而生，因為網路的流行而成為相當熱門的研究課題。

另外，視訊透過網路傳遞很方便，但是在傳輸的過程中經常發生封包遺失的情形。在wireless環境中，由於視訊的傳輸經常會遭遇雜訊干擾與通訊頻道擁塞等問題，導致傳輸的視訊收看品質會遽烈降低。
在本論文我們提出因應視訊在網路傳輸面臨問題的解決方法。主要目的有二：一、對視訊做版權的保護。二、對視訊發展強健性傳輸技術。

在視訊的版權技術方面，我們提出一個強健性的視訊浮水印技術，滿足浮水印透明性、即時性的浮水印偵測、bit-rate 保持不變、在壓縮域(VLC domain)的處理等等的要求。為滿足浮水印即時性的偵測的需求,在崁入時，我們的方法是直接在VLC 域上作處理，因此在作浮水印偵測時，也是直接在VLC域作即時性的偵測。另外為了處理 copy attack 與 collusion attack, 我們提出 video frame-dependent watermark (VFDW)方法，對於copy attack 與 collusion attack 具有良好的抵抗能力。

在視訊強健性傳輸技術方面，我們提出一個以hash為基礎的強健性的視訊傳輸技術。我們提出hash定義,此hash用來描述被參考的區塊(block)。當視訊在作壓縮時，把被參考的區塊的hash崁入視訊中。一旦視訊的封包遺失，則使用者端可將原先藏在視訊中的hash取出，利用hash來幫助作視訊的修補。此外, 我們也提出2階段式的hash比對方法,與結合hash資訊的side-match方法。從實驗上來看，我們的方法在burst packet lose的環境中具有良好的抵抗能力。

關鍵字

浮水印 ； 版權保護 ； 強健性傳輸 ； 資料隱藏

並列摘要

Due to the rapid development of network in the past decade, a large number of multimedia data have been stored and transmitted via network.
Digital multimedia is much superior to analog data because it can easily copy without losing any quality.
However, this superiority is also a drawback because digital multimedia could be easily tampered with, duplicated or distributed.
As a consequence, the intellectual property protection problem has become an urgent issue in the digital world.

In addition, it is very easy and convenient to transmit the digital video over network.
However, video transmission usually suffers interference and bandwidth congestion seriously under wireless network environment.
The reconstructed quality will degrade drastically.
Therefore, robust video transmission becomes an important issue.

In this thesis, we proposed two research topics including video copyright protection and robust video transmission.
For video copyright protection, we proposed a robust video watermarking scheme.
Some video watermarking issues, including compressed domain watermarking, real-time detection, bit-rate control, and resistance to watermark estimation attacks, will be satisfied.
In the embedding process, our algorithm is designed to operate directly in the variable length codeword (VLC) domain.
In addition, the watermark detection is also to operate directly in the variable length codeword (VLC) domain to satisfy the requirement of real-time detection.
For bit-rate control, we describe how watermark signals can be embedded into compressed video while keeping the desired bit-rate nearly unchanged.
In particular, in order to deal with both collusion and copy attacks that are fatal to video watermarking, the video frame-dependent watermark (VFDW) is presented.
From the experiment result, our video watermarking scheme is robust to collusion and copy attack, simultaneously.

For video transmission, error concealment is adopted at decoder when video data is corrupted or lost.
In general, error concealment exploits the smoothness of the corrupted block boundary for data recovery.
It is a kind of side-matching.
In other words, side-matching is dominanted by the boundary smoothness instead of the block content.
We propose a data hiding-based hash matching scheme to estimate the lose motion vector for concealment at decoder.
The proposed video block hash is used to describe the reference block.
In the encoder side, the hash of each reference block is hidden into video itself.
Therefore, decoder can extract the hash for error concealment when packets are lost.
Furthermore, we propose the two-stage hash matching procedure for searching the content-based candidate blocks.
In addition, the extracted hash can be adopted in side-matching process for finding the final candidate block.
From the experimental results, the video quality is improved under the burst packet lose environment.

並列關鍵字

copyright protection ； watermarking ； data hiding ； error resiliece

參考文獻

[5] “Draft ITU-T recommendation and final draft international standard of joint video

Compensation at the Receiver,” Proc. IEEE Int. Conf. on Image Processing,

[8] I. Agi and L. Gong, “An empirical study of secure MPEG video transmissions,”

Proc. Internet Society Symp. Network and Distributed System Security, pp. 137-

144, 1996.