影響偽裝影像失真程度因素分析

Analysis of Factors Influencing Distortion of Stego-image

呂侑靜(YU-CHING LU)

指導教授 ： 呂慈純

朝陽科技大學/資訊學院/資訊管理系/碩士(2018年)

摘要

近年來，資訊隱藏技術的應用越來越廣泛，許多學者為了提高影像品質或是增加資訊負載量，提出了許多創新的方法。再者，影響影像品質的原因眾多，大部分的研究以降低修改量為目標，以達到高影像品質。本論文發現機密資料的大小，亦是影響影像品質的重要因素之一，除此之外，藏入時，影像的區塊大小，也會影響像素間的相關程度，間接影響藏入效能。因此，本論文以機密資料以及區塊大小為研究主軸，探討可能影響影像品質的因素。本論文將分為兩個部分，第一部分以改良機密資料為主，提出兩個不同的方法，第一個方法以中間對折策略結合像素值排序法，透過中間對折策略縮小機密資料，以中間對折策略所產生的偽裝影像，擁有較佳的相似度，且能提升複雜影像的相關程度。因此，第二階段再以像素值排序法來進行二次藏入；第二個方法根據機密資料出現的頻率進行重新編碼，出現機率較高的機密資料，給予較短的編碼，如此一來，不會有資料冗餘的問題發生，接著以新的編碼進行藏入。第二部分以不同的區塊大小進行實驗，改良Gulve 和 Joshi學者的方法，以1×3、2×2、2×3、3×3區塊，根據不同的區塊大小以及不同的影像特性進行分析，觀察其藏量及影像品質的變化，經實驗結果發現，於2×2的區塊中能有較好的藏入效能。

關鍵字

機密符號 ； 雙影像技術 ； 像素排序法 ； 龜殼藏入法 ； 中間對折

並列摘要

In recnet years, information hiding technology is widely useful. Many researchers proposed the novel schemes to enhance the quality of image and capacity. However, there are many factors which cause image distorstion. How to make the stego-image closer to the original image is an important issue. Many studys try to reduce the distance between the original pixels and stego pixels or narrow the modification scale of the pixel to enhance the image quality of stego-image. In this study, we obsevered that the secret data is a key factor to influence the image distorstion. In addition, the size of block also effect the correlation between the neighbor pixels. In this manner, the hidden performance will be influenced. Thus, this study is divided into two parts. First part of this study is focused at reducing the secret data. Two novel methods are proposed. The first method combines the center folding strategy with pixel-based pixel value ordering (PPVO). All the secret data is folded by Lu et al’s center folding method. The dual images which generated with the center folding strategy are very similar to the original image. Thus, the stego-images are suitable for the second times hiding. The second part of this study, the digital image is used to be the secret data. In the fisrt step, the secret image is transferred into a binary format and the method calculates the frequency of the occurrence of each binary value. In the hiding procdure, the same hiding rule as Chang et al’s method is used. In the second part, we focus on the size of the block. The realtion between the neighbor pixels within block has higher reatlatd. Therefore, this study tries to improve Gulve and Joshi’s method by adjusting the size of the block to several different sizes, including 1×3, 2×2, and 3×3, and then determines the effectiveness of the resulting embedding capacities and image qualities of the stego-images. Based on the image features, we determine the appropriate block size for different images. According to the aforementioned, the block size is set to be 2×2 could obtain a better PSNR value.

並列關鍵字

Secret Symbol ； Dual Image ； Pixel Value Ordering ； Turtle Shell ； Center Folding

參考文獻

[1] 呂慈純、陸哲明、張真誠，多媒體安全技術，全華圖書股份有限公司，2007。

[2] A.M. Alattar(2004), “Reversible Watermark Using the Difference Expansion of a Generalized Integer Transform,” IEEE Transactions on Image Processing, vol. 13, no. 8, pp. 1147-1156.

[3] A.K. Gulve, and M.S. Joshi(2015), “A High Capacity Secured Image Steganography Method with Five Pixel Pair Differencing and LSB Substitution,” Graphics and Signal Processing, vol. 7, no. 5, pp. 64-74.

[4] B.Qu, X. L.Li, Y.Zhao, and R.R. Ni(2014), “Reversible Data Hiding Using Invariant Pixel-Value-Ordering and Prediction-Error Expansion,” Signal Processing, vol. 29, no. 7, pp. 760-772.

[5] C.K. Chan, and L.M. Cheng(2004), “Hiding Data in Images by Simple LSB Substitution,” Pattern Recognition, vol. 37, no. 3, pp. 469-474.