利用權重模數技術的資訊隱藏演算法

本文提出一個利用權重模數（weighted modulus）技術的高藏量高品質灰階影像資訊隱藏演算法，簡稱WM方法。我們先使用理論分析來求出在不同像素位元嵌入量下，偽裝影像具有的理論最小變動量。接著，我們提出如何在不同像素位元嵌入量下，尋找最佳權重的方法。使用最佳權重來嵌入秘密訊息時，使符合或儘可能的相近於理論最小變動量。我們的演算法使用此最佳權重來嵌入秘密訊息，產生高視覺品質的偽裝影像。理論分析結果顯示：在相同的嵌入量下，WM演算法可提高影像的PSNR（peak signal to noise ratio）數值；實驗結果也佐證演算法產生之偽裝影像品質優於目前文獻上現有的高藏量高品質資訊偽裝演算法。WM方法周延可以使用多個像素來嵌入密訊息，同時也具有高度彈性，可以使用線性內插滿足使用者需求。演算法具有抵抗RS偽裝偵測攻擊之安全性。本文的具體貢獻為提出一個WM演算法。該資訊隱藏演算法具有高藏量低失真、周延、彈性、安全等特性。實驗結果證實WM演算法優於我們所知目前文獻最新的資訊隱藏演算法。WM演算法擴展影像偽裝學的應用範疇。

關鍵字

資訊隱藏；權重模數技術；灰階影像；高藏量與高視覺品質；偽裝學

並列摘要

This paper proposes a data hiding algorithm using a weighted modulus technique for grayscale images. We analyze the pixel variation incurred by the message embedding, and we describe an approach to determining the optimal/effective weights that produces the stego image with as small pixel variation as possible. Once a desired capacity is given by users, our algorithm employs the optimal/effective weights to embed secret message into the cover image. This allows us to produce a stego image with minimal/ less pixel variation and high visual quality. In comparison with our counterpart, the theoretical analysis shows that our scheme produce higher PSNR (Peak Signal to Noise Ratio) value when embedding the same amount of the secret message. Experimental results verify that our scheme has better performance than our counterparts. Our scheme can use many pixels as a group to embed secret message, and it can use a linear combination approach to satisfy the user different demand and is secure against the RS steganalysis attack. In conclusion, the major contribution of our algorithm is to provide high capacity and low distortion characteristic to conceal secret message for grayscale images, extending the image-based steganographic applications.

並列關鍵字

data hiding ； weighted modulus technique ； grayscale image ； high capacity and high quality ； steganography

國際替代計量

利用權重模數技術的資訊隱藏演算法

全文下載

主題瀏覽