Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/5993
標題: 基於展頻技術之數位浮水印
Digital Watermarking Based on Spread Spectrum Technique
作者: 黃恆傑
關鍵字: 數位浮水印;離散餘弦轉換;展頻
出版社: 電機工程學系
摘要: 
電腦與通訊網路的進步,使得電子文件交換、電子商務、隨選視訊、數位圖書館等數位服務蔚為風尚。但是,由於數位化資料具有容易且精確地被複製的特性,所以需要有效的保護機制來防止竊取或篡改,數位浮水印(digital watermarking)就是產學界高度重視的技術之一。數位浮水印技術為近年來新興的技術,它能將資訊隱藏於多媒體資料中用來宣告所有權、著作權或認證。
在我們的研究上,我們是使用通訊系統中的展頻通訊原理來實現數位浮水印。本文將數位浮水印系統比擬成通訊系統,原始媒體資訊(影像)之頻率域為傳輸通道,而浮水印資料為傳輸訊號,可能遇到的攻擊或破壞則視為通道雜訊[1][23]。展頻通訊已區分為直接序列展頻(Direct Sequence Spread Spectrum DSSS)與跳頻展頻(Frequency Hopping Spread Spectrum FHSS)二種,因此,我們針對這二種展頻技術而設計出跳頻展頻浮水印(FHSS Watermarking)與直接序列展頻浮水印(DSSS Watermarking)。
JPEG壓縮是現今普遍使用的壓縮格式,因它有著高壓縮特性,以及壓縮後影像品質仍很高,所以普遍被大眾使用。在本論文中,我們針對JPEG壓縮方式提出其數位浮水印的嵌入構想,將浮水印嵌入於對JPEG量化有較強抵抗力的頻帶中,以達到提高抵抗JPEG壓縮之目的。由於JPEG是使用離散餘弦轉換將影像轉換至頻率域,為了配合其特性,所以我們的展頻數位浮水印在轉換領域部分也是基於離散餘弦轉換。
實驗結果顯示,我們所提出的跳頻展頻浮水印在維持影像品質的情形下,對於JPEG壓縮具有高度的強韌性,另外使用PhotoImpact 5影像處理套裝軟體來測試比較跳頻展頻浮水印與直序展頻浮水印對各種影像處理特效其抵抗破壞效能的優劣,我們發現跳頻展頻浮水印對幾何失真破壞比較有抵抗力,而直序展頻浮水印則是訊號處理破壞具有抵抗力。
本論文最後提出融合跳頻展頻與直序展頻的Combined SS Watermarking技術。Combined SS Watermarking可同時擁有跳頻展頻浮水印與直序展頻浮水印的優點,如此,Combined SS Watermarking技術便可以抵抗較多種的破壞,達到提高數位浮水印強韌度之目的。

The advances of computers and communication networks make digital content service such as digital document exchange, E-commerce, video on demand, and digital library very popular. Since digital data can be easily reproduced without any loss, copyright protection becomes an imperative requirement to prevent piracy. Digital watermarking has been proposed as one of the techniques to prevent data piracy and plagiarism. Digital watermarking is a novel and emerging technology. It has developed very quickly for the past few years. A digital watermark is a set of information that is robustly and imperceptibly embedded in the data to be protected. Applications include copyright protection, and authentication.
In the thesis, we developed two watermarking techniques using spread-spectrum technique. We interpreted the watermarking technology based on communication theory. The watermark was the signal to be transmitted. The frequency domain of the original image was the transmission channel. The attacks were regarded as channel noise. There were two methods to implement the spread-spectrum system: direct- sequence spread- spectrum (DSSS) and frequency-hopping spread-spectrum (FHSS). Based on these tow methods, we develop FHSS Watermarking and DSSS Watermarking.
JPEG compression is the most widely used image because compression technique of its high compression ratio and reasonable image quality. Here, we develop the watermark embedding method especially for JPEG encoding process since block Discrete Cosine Transform (DCT) was used for JPEG. In our approach, a block DCT-based algorithm was used to embed the image watermarking. In order to survive the JPEG compression, the watermark must be embedded in the bands that can with stand the JPEG quantization.
Experiments showed that FHSS Watermarking is more resistant to the attack of JPEG compression. The PhotoImpact 5 digital image processing software was used for other attacks. We found that FHSS Watermarking is more robust than DSSS Watermarking in geometric distortions. But in other signal processing modifications, DSSS Watermarking is more robust than FHSS Watermarking. Therefore, we developed a new method that combined with FHSS Watermarking and DSSS Watermarking, we call the new method “combined SS Watermarking”. The experimental results prove that combined SS Watermarking has both the advantages of FHSS Watermarking and DSSS Watermarking. Therefore, combined SS Watermarking is more robust and can with stand many different attacks.
URI: http://hdl.handle.net/11455/5993
Appears in Collections:電機工程學系所

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.