基于數(shù)字水印的圖像認證技術研究
發(fā)布時間:2018-07-18 14:20
【摘要】:數(shù)字水印是一種有效的數(shù)字產(chǎn)品版權保護和數(shù)據(jù)安全維護技術,是信息隱藏技術研究領域的一個重要分支。采用數(shù)字水印技術進行圖像認證,具有與數(shù)字簽名不同的優(yōu)勢,它已經(jīng)成為當前信息認證領域研究的熱點。本文比較系統(tǒng)地研究了數(shù)字水印的重要分支一認證水印(脆弱性數(shù)字水印和半脆弱性數(shù)字水印)的發(fā)展情況:論述了認證水印提出的背景、基本框架;重點分析了脆弱性數(shù)字水印和半脆弱性數(shù)字水印現(xiàn)有的各種算法,比較了它們的優(yōu)缺點,在此基礎上提出了三種認證方案。 在方案一中,本文提出了一種新的摘要提取方法,通過理論分析及上萬次的仿真,證明:該方法可以獲得較短的摘要長度,并且當輸入發(fā)生變化時,該方法產(chǎn)生的摘要比采用Hash函數(shù)奇偶校驗位和Hash-hamming重量的方法具有更高的變化概率。在此基礎上,提出了一種新的脆弱性數(shù)字水印算法,理論及實驗結果證明該方法具有很高的圖像篡改檢測概率;較高的篡改定位精度;同時可以較好的保持圖像的使用價值;算法簡單、安全性高,具有較好的實用性。 在方案二中,本文將方案一的方法進一步擴展到半脆弱性數(shù)字水印中。首先提出了一種新的圖像預處理方法,該方法可以使圖像具有很好的抗JPEG壓縮能力。其次,本文在應用預量化原理的基礎上,將提取出的水印信息嵌入到DCT交流系數(shù)中,使得圖像在壓縮后水印信息可以完全重構。該方案不僅可以區(qū)分正常的壓縮和篡改,而且可以進行篡改準確定位,同時具有很高的篡改檢測概率。方案的實現(xiàn)算法簡單,安全性高。仿真實驗結果證明了方案的有效性。另外,在方案中本文還提出了一種新的篡改定位方法—區(qū)域定位和交叉定位復合的方法,這種方法可以有效減少誤判。 在以上研究的基礎上,本文對DWT域的半脆弱水印也進行了相關研究。
[Abstract]:Digital watermarking is an effective digital product copyright protection and data security maintenance technology, is an important branch of information hiding technology. Image authentication based on digital watermarking has different advantages from digital signature, and it has become a hot topic in the field of information authentication. In this paper, the development of authentication watermarking (fragile digital watermarking and semi-fragile digital watermarking), an important branch of digital watermarking, is systematically studied. The background and basic framework of authentication watermarking are discussed. The existing algorithms of fragile digital watermarking and semi-fragile digital watermarking are analyzed, their advantages and disadvantages are compared, and three authentication schemes are proposed. In the first scheme, a new abstract extraction method is proposed. Through theoretical analysis and thousands of simulations, it is proved that the method can obtain a shorter length of abstract, and when the input changes, The method has a higher probability of variation than that of using Hash function parity check bits and Hash-hamming weights. On this basis, a new fragile digital watermarking algorithm is proposed. The theoretical and experimental results show that this method has high image tamper detection probability, high tamper location accuracy, and can maintain the value of image. The algorithm is simple, safe and practical. In the second scheme, the method of scheme one is further extended to semi-fragile digital watermarking. Firstly, a new image preprocessing method is proposed, which can make the image have a good ability to resist JPEG compression. Secondly, based on the principle of prequantization, the extracted watermark information is embedded into the AC coefficients of DCT, so that the watermark information can be reconstructed completely after compression. This scheme can not only distinguish the normal compression from the tamper, but also can locate the tamper accurately, and has a high probability of tampering detection. The algorithm of the scheme is simple and the security is high. The simulation results show that the scheme is effective. In addition, a new tamper localization method, the combination of region location and cross location, is proposed in this paper, which can effectively reduce misjudgment. Based on the above research, this paper also studies the semi-fragile watermark in DWT domain.
【學位授予單位】:西北大學
【學位級別】:碩士
【學位授予年份】:2006
【分類號】:TP309.7
本文編號:2132170
[Abstract]:Digital watermarking is an effective digital product copyright protection and data security maintenance technology, is an important branch of information hiding technology. Image authentication based on digital watermarking has different advantages from digital signature, and it has become a hot topic in the field of information authentication. In this paper, the development of authentication watermarking (fragile digital watermarking and semi-fragile digital watermarking), an important branch of digital watermarking, is systematically studied. The background and basic framework of authentication watermarking are discussed. The existing algorithms of fragile digital watermarking and semi-fragile digital watermarking are analyzed, their advantages and disadvantages are compared, and three authentication schemes are proposed. In the first scheme, a new abstract extraction method is proposed. Through theoretical analysis and thousands of simulations, it is proved that the method can obtain a shorter length of abstract, and when the input changes, The method has a higher probability of variation than that of using Hash function parity check bits and Hash-hamming weights. On this basis, a new fragile digital watermarking algorithm is proposed. The theoretical and experimental results show that this method has high image tamper detection probability, high tamper location accuracy, and can maintain the value of image. The algorithm is simple, safe and practical. In the second scheme, the method of scheme one is further extended to semi-fragile digital watermarking. Firstly, a new image preprocessing method is proposed, which can make the image have a good ability to resist JPEG compression. Secondly, based on the principle of prequantization, the extracted watermark information is embedded into the AC coefficients of DCT, so that the watermark information can be reconstructed completely after compression. This scheme can not only distinguish the normal compression from the tamper, but also can locate the tamper accurately, and has a high probability of tampering detection. The algorithm of the scheme is simple and the security is high. The simulation results show that the scheme is effective. In addition, a new tamper localization method, the combination of region location and cross location, is proposed in this paper, which can effectively reduce misjudgment. Based on the above research, this paper also studies the semi-fragile watermark in DWT domain.
【學位授予單位】:西北大學
【學位級別】:碩士
【學位授予年份】:2006
【分類號】:TP309.7
【引證文獻】
中國碩士學位論文全文數(shù)據(jù)庫 前1條
1 董萱;基于脆弱水印的圖像認證技術研究[D];西安電子科技大學;2011年
,本文編號:2132170
本文鏈接:http://www.sikaile.net/kejilunwen/ruanjiangongchenglunwen/2132170.html
最近更新
教材專著