論文摘要

AbstractImage hiding aims to embed a secret image into a cover image to form astego-image, transmitted over the public networks without revealing the secret image.To our best knowledge, the previous research of image hiding mainly focuses onprivacy, good image quality, and high hiding capacity. However, there are twoinevitable issues, not address in present image hiding schemes. Firstly,authentication, ascertaining the origin and integrity of the transmitted images, is animportant issue for the transmitted images. Secondly, some parameters of imagehiding are all shared by sender and receiver in advance so that the hidden image canbe correctly extracted. Nevertheless, if a lot of information has to beshared/transmitted by a secure channel in advance, image hiding becomes impractical.In this thesis, a new secure image hiding scheme, in which the VQ technique isadopted to compress secret images before embedding, is proposed to convey severalsecret images. Other then satisfying the basic properties of image hiding such asprivacy, good image quality, and high hiding capacity, the proposed scheme has othervaluable advantages. Firstly, for transmitting codebook, not only secret images butalso the corresponding codebook is embedded into the cover image. As a result, thecodebook can be automatically extracted by the intended receiver for decompressionin extracting process. Secondly, the proposed scheme highlights a future, imageauthentication of both the stego-image and secret images, often existing in imagesecurity mechanisms but seldom addressed in present image hiding schemes.Thirdly, for transmitting the parameters used in extraction process, digital signature isalso applied to convey the corresponding parameters.

************************************************************************************

影像隱藏是將一張祕密影像藏到一張比較不重要的掩護影像中，然後產生一張內含祕密影像的偽裝影像。利用這樣的技術可以將內含祕密影像的偽裝影像透過網路安全地傳送到對方，而不會洩露出其中內含祕密影像。就我們所知，在先前的影像隱藏研究文獻中，主要都集中在影像的隱私性、良好的影像品質以及高隱藏容量的議題上。然而，有二個重要且必須的影像安全議題，但在先前的影像隱藏研究中，卻沒有人把它們納入到影像隱藏的機制上。第一個是驗証，確保被傳送影像的資料來源正確及它的完整性,這對被傳送的影像來說是很重要的議題。第二個是一些用來取出祕密影像的祕密參數，通常由傳送方和接收方事先溝通好，以便能正確地取出祕密影像。但是，如果太多的的祕密資訊在傳送祕密影像前就要事先溝通好，那這樣的影像隱藏機制就變得不實用。在這篇論文中，提出一個新的影像隱藏機制，它可以用來傳送多張的祕密影像。但是，在將祕密影像藏入掩護影像之前，必須先用向量量化的技術將祕密影像做壓縮以縮小隱藏資訊的量。另外，在這機制中，不僅滿足資訊隱藏的基本屬性如影像的隱私性、良好的影像品質以及高的隱藏容量。它同時還具有以下的加值優點。第一個優點是，我們不只隱藏了祕密影像，同時也將向量量化編碼簿藏到掩護影像裏。而且，在使用者要取出祕密影像時，向量量化編碼簿可以自動的取出做解壓縮的動作。第二個優點是，我們提出一個不同的觀點。就是影像驗証，同時對傳送的偽裝影像和祕密影像做影像認証。第三個優點是，將那些用在取出祕密影像的祕密參數加在數位簽章中，利用數位簽章將這些參數傳送給接收者。因此，數位簽章不僅提供了驗証傳送影像的功能，同時也用來傳送相關的祕密參數。