摘 要:針對電子文檔真實性和原始性的識別問題,本文提出了在DCT域中采用抖動調制直流和低頻系數添加水印的算法。首先,對載體圖像預處理,并將標志圖像轉換為容量較小的QR碼;其次,對載體圖像做分塊DCT變換;接著,采用抖動調制方式將水印嵌入分塊DCT系數的直流、低頻部分;最后,對嵌入水印的載體圖像做DCT反變換。通過仿真實驗,對縮放、裁剪、高斯噪聲、壓縮等攻擊后的文檔中的水印進行了提取。實驗表明,本算法具有水印隱藏效果好,能夠經受一定比例縮放、一定量的圖像裁剪和適度的JPEG壓縮等攻擊,能夠實現盲檢測及嵌入較大容量標志信息的優點。
關鍵詞:數字水印;二值圖像;離散余弦變換;快速響應矩陣碼;抖動調制
中圖分類號:TP309.7 文獻標識碼:A 文章編號:2096-4706(2018)12-0013-04
Digital Watermarking Algorithm of QR Code Based on DCT Domain
LIU Xiaohong,GONG Ruichun
(School of Information Science and Technology,Baotou Teachers’College,Baotou 014030,China)
Abstract:Aiming at the identification of authenticity and originality of electronic documents,this paper proposes a method of adding digital watermark in DCT domain by using dither modulation DC and low frequency coefficients. Firstly,the carrier image is preprocessed,the logo image is converted to a smaller capacity of QR code;Secondly,the carrier image is divided into blocks,DCT transform;then,the watermark is embedded into the DC and low frequency coefficients of the block DCT by using the dither modulation method. Finally,the DCT inverse transform of the carrier image containing the watermark is done. The watermark information is extracted from the electronic documents after the attack,such as scale,cropping,Gauss noise,compression and so on. The results show that this algorithm has the advantages of good watermark hiding effect,it is able to withstand a certain scaling,a certain amount of image cropping and moderate JPEG compression coding etc.,and blind detection of watermark information and embedded larger logo image.
Keywords:digital watermarking;two value image;discrete cosine transform;fast response matrix code;dither modulation
0 引 言
隨著數字化技術的迅猛發展,電子信息的存儲、傳遞遍及各個領域。電子文檔正在逐漸取代傳統文檔。然而,由于電子文檔具有流動性、易復制性、信息的不穩定性,因而,無法從外觀上直接辨別真偽,使得電子文檔的真實性和原始性容易受到質疑。
數字水印通過在數字媒體信息中添加標志信息,保護數字媒體的版權。將水印信息嵌入載體文件,不影響原文件的外觀和完整性。這樣,既不影響原載體的使用價值,又不容易被探知和再次修改,但可以被生產方識別和辨認。通過這些隱藏在載體中的信息,可以達到確認內容創建者、傳送隱秘信息或者判斷載體是否被篡改等目的[1]。……