基於差值擴張方法之可視可逆浮水印技術研究

機構典藏 > College of Engineering > Graduate Institute & Department of Computer Science and Information Engineering > Thesis > Item 987654321/111359

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/111359

Title:	基於差值擴張方法之可視可逆浮水印技術研究
Other Titles:	Difference-expansion based reversible and visible watermarking scheme
Authors:	葉信呈;Yeh, Hsin-Cheng
Contributors:	淡江大學資訊工程學系碩士班陳建彰;Chen, Chien-Cheng
Keywords:	可視浮水印;可逆浮水印;差值擴張;動態像素值配對;reversible watermark;visible watermark;difference expansion;dynamic pixel value mapping
Date:	2016
Issue Date:	2017-08-24 23:50:50 (UTC+8)
Abstract:	可視可逆浮水印技術可以將浮水印資訊嵌入至數位影像中以獲得浮水印嵌入影像，可以在浮水印嵌入影像中明顯發現可辨識之浮水印資訊，使數位影像的版權受到保護。當需要時可以由浮水印嵌入影像中抽取浮水印資訊，並還原獲得原數位影像。本論文提出一個具顯著嵌入效果之可視可逆浮水印技術，此技術使用動態像素值配對技術將浮水印資訊嵌入至數位影像中選定的可視浮水印區域R，並使用差值擴張法將相關二值資訊嵌入至整張影像當中。可視浮水印區域的大小將受到區塊設定值k及二值浮水印影像邊長的影響而改變，當選擇的k值越大我們將得到較大的可視浮水印區域，而k值的範圍介於1至數位影像與浮水印影像的最小邊長比。首先依據所選擇的k值將可視浮水印區域分割為多個大小為k×k的不重疊區塊，每個區塊對應一個浮水印資訊，並使用本論文提出的動態像素值配對方法嵌入高視覺效果的可視浮水印。接著使用金鑰將二值浮水印資訊及浮水印資訊為黑色像素所對應之k×k區塊之LSB位元陣列進行加密後，使用差值擴張方法將二值資訊嵌入至數位影像中。實驗結果顯示我們能夠在浮水印嵌入影像中，清楚看到明顯的可視浮水印資訊，使數位影像能夠明確得以表示影像的所有權。而在需要抽取浮水印時，使用正確金鑰即能夠有效得將浮水印資訊由浮水印嵌入影像中完整抽取，並無失真得還原原始的數位影像。 A visible and reversible image watermarking scheme is a technique for embedding watermark into digital image, after embedding watermark, the visible watermark is visibly and clearly shown on the embedded watermarking image for protecting the ownership of the digital image. The watermark can be extracted with recovering the original digital image when needed. This paper presents a visible and reversible image watermarking scheme, using the proposed dynamic pixel value mapping to adjust part of the digital image, named visible embedded region R. The size of R is determined by the block coefficient k and also the size of the binary watermark image, in which large k value leads to large visible embedded region, then using the difference-expansion method to embed required binary strings into the whole image. The visible embedded region R is partition into non-overlapped k×k blocks and each block is related to one bit of the watermark image. For those blocks that are related to the black pixels of the watermark image, these k×k blocks are adjusted by the dynamic pixel value mapping for highly visual detection. Using the secret key to encrypt the binary bit string S, which composed of the binary watermark image and LSB bits of the black pixels corresponded k×k blocks, is embedded into the digital image using the difference-expansion method. Experimental results show that the watermark information is clearly embedded into the embedded region R and distortion of the reversible embedding is limited.
Appears in Collections:	[Graduate Institute & Department of Computer Science and Information Engineering] Thesis

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	191	View/Open

Loading...