利用圖片攪亂以加強浮水印之安全性

Abstract

在設計浮水印系統時，必須考慮幾項主要的條件和需求：不易察覺、強韌性、明確性、安全性和容量。因為這些條件之間會有取捨上的兩難（例如，增加容量可以提高強韌性，卻可能降低隱密性），多數浮水印的設計都會盡可能取得一個適當的平衡點，以達到浮水印演算法整體的最佳效能。 本文以WTQ演算法為論述起點，探討該演算法的設計架構，以及其他學者針對此演算法所做的攻擊，延伸討論浮水印的安全性。為了增強浮水印的安全性，本文引用圖片攪亂的方法，作為浮水印演算法的前置處理，以此加強浮水印的安全性。 本研究中，將比較幾種同樣採用小波樹（wavelet tree）架構所設計的浮水印演算法（WTQ、WTGM和DEED），觀察這些演算法在圖片攪亂的輔助下所產生的變化，並利用統計特徵值進行原圖與攪亂之圖片的分析，以說明圖片攪亂對浮水印強韌性造成影響的原因。

When designing a watermarking algorithm, we should take account of some requirements, such as unobtrusiveness, robustness, unambiguousness, security and capacity. There are some trade-offs between these requirements, and the total performance of watermarking system would decline if some requirements do not been taken into account carefully. In the paper, we first introduce the WTQ scheme, and then explain how this watermarking algorithm can be attacked by cryptanalysis. Therefore, we review the “security” requirement of WTQ, and use image scrambling as a pre-processing mechanism which can improve the security of WTQ. After confirming the useful outcome of image scrambling on the security of WTQ, we also need to review the influence of image scrambling on the robustness of WTQ. Therefore, we compare the performance of three wavelet-tree-based watermarking schemes with and without using image scrambling, and explain the results by the analysis of statistical features.