| 摘要: | 摘要
在資訊發達的數位時代，隨著網際網路的普及，任何的數位資訊都能透過網際網路廣泛地傳播與散佈，然而享受數位科技帶來的便捷與商機時，有些潛在的問題，比如隱私權及智慧財產權侵犯，著作物的違法複製、歪曲、偽造、割裂或篡改，資料竊取、竊聽與偷窺等，因此，要如何保護數位資料在網路上傳輸時的安全，便成為現今數位時代不容忽視的重要議題。
近年來，資訊安全技術不斷的被學者所提出，例如：密碼學、機密共享及資訊隱藏等，各種資訊安全皆有不同的優缺點。例如，傳送方將數位資料以金鑰加密產生密文，密文透過網路傳送到接收方，再以相同的金鑰做解密，以取得機密資料。且因駭客沒有金鑰可以解密，但駭客可以攔截並破壞機密資料，再將已破壞的機密資料傳送給接收方，如此接收方即使有金鑰也無法還原原始機密。
為了解決這樣的問題，已有眾多學者提出資訊隱藏的概念，當資訊隱藏在嵌入更多的機密數據及數值較大的情況下，本文提出了運用差值編碼實現高品質偽裝影像資訊隱藏方法。該方法使用局部特徵和全域特徵，鄰近機密數據將大多數機密數據轉換為較小的值。反之，一些機密數據被轉換為較大的值。由於嵌入較小的值不會引起嚴重的圖像失真問題，因此，在具有相同嵌入率的情況下，該方法的圖像質量高於以前的方法。
關鍵詞：資訊隱藏、動能絕對值區塊截短碼、PSNR值、差異擴張法。
;Abstract
Any digital information can be transmitted widely by internet as Internet become universal. It implies that the digital techniques make that life become convenience. However, there are some potential issues, e.g., private, copyright protection, information copy, tampering, and so on. Therefore, how to protect the digital data transmitted on Internet becomes an important issue.
Recently, information security techniques have been proposed continually, e.g., cryptography, secret sharing, and data hiding. Each information security technique owns different advantages and disadvantages. For example, a sender used key to encrypt digit information as cipher. After transmitting cipher to receiver, `the information can be decrypt to obtain secret data. Although hackers do not have key to decrypted data, he/she can intercept or destroy data, making that receiver cannot obtain correct information.
In order to solve this problem, many scholars proposed the concepts of data hiding. In this paper, a difference encode method is proposed to embed more secret data, where the secret data is large. The proposed method used local and global characteristic to transform most secret data into the smaller values. On contrary, a few secret data are converted into the larger values. Since embedding the smaller value does not invoke serious image distortion, the proposed method has better image quality than the previous method with the same embedding ratio.
Keyword: Data hiding, Absolute moment block truncation coding, PSNR, difference expansion |
