應用於智慧財產權保護與認證之數位影像浮水印

Abstract

近年來，數位化科技的進步與網際網路的發達使得資料存在的形式逐漸由實體化轉成數位化。數位化後的內容具有無失真快速傳播與複製的優點。然而，如刀之兩刃，數位科技也同時帶給仿冒者更多的機會剽竊原作者的創意與所有權。為了解決這個問題，在這篇論文中我們利用數位浮水印技術保護數位影像的智慧財產權以及偵測影像內涵是否遭惡意竄改。由於數位浮水印技術可分為強健式浮水印、易碎式浮水印及標題式浮水印。針對第一類浮水印技術，我們提出一個方法解決影像之智慧財產權保護與影像內涵認證之問題。在此篇論文中，我們所提出的方法是採用gray-level的影像來當浮水印，並利用離散餘絃轉換，量化和展頻的技巧來達到浮水印系統的強韌性。我們利用離散餘絃轉換使得浮水印較不易受JPEG的影響，利用量化和展頻的技巧來降低藏入的資料量，將浮水印藏入原圖中頻的部分，實驗的結果品質大致良好，此方法在對抗一些浮水印攻擊如濾波、JPEG壓縮、剪裁等，都有不錯的效果。

The advances of digital techniques and Internet have promoted the style of contents from physical form to digital form. Digitized contents can be distributed and duplicated rapidly without any degradation. This property is both an advantage and a drawback of digital technique. Pirate may distribute or modify digitized contents without the permission of the creator or owner. This behavior will be of no doubt a threat to daily digital life. The objective of this dissertation is to investigate how to protect the ownership and contents of digital images using the technique of digital watermarking. According to different purposes, watermark technology can be categorized into robust, fragile, and captioning watermarking. In this paper, we focus on the robust watermarking. We will propose techniques to solve the problem of copyright protection, content authentication. The proposed watermarking scheme adopts a gray-level image as the watermark. Besides, discrete cosine transformation (DCT) technique and quantization method are applied to strengthen the robustness of the watermarking system. Both original image and digital watermark, processed by DCT transformation, can build a quantization table to reduce the information size of the digital watermark. After quantized watermark is embedded into the middle frequency bands of the transformed original image, the quality of the watermarked image is always visually acceptable because of the effectiveness of the quantization technique. The experimental results show that the embedded watermark can resist image cropping and JPEG lossy compression.