植基於邊緣偵測及最佳像素調整之資訊隱藏方法

An Improved Steganographic Method Based on Median Edge Detection and Optimal Pixel Adjustment Process

劉興漢(Liu, Hsing-Han) ； 賀盛志(Ho, Sheng-Chih) ； 鄭岳原(Cheng, Yue-Yuan)

《國防管理學報》 45卷1期 (2024/05) Pp. 23-44

https://doi.org/10.29496/JNDM.202405_45(1).0002

摘要

本研究提出植基於邊緣偵測及最佳像素調整之資訊隱藏方法，首先利用中間邊緣偵測法用來區分目標像素是位於邊緣或是平滑區，據以判定秘密訊息的藏入量，基於人類視覺系統，若目標像素位於邊緣則以最低有效位元取代與最佳像素調整法嵌入較多的秘密訊息，若在平滑區則以相同方法嵌入較少的秘密訊息。實驗中，我們使用Lena等8張灰階影像進行測試，得到高藏密量且偽裝影像品質良好的效果。此項方法不需複雜的演算過程，透過現行基本的軟硬體設備，即可將機密資訊藏入影像中，並從偽裝影像取出隱藏訊息。此研究可提供國防資訊安全管理研究領域參考應用。

關鍵字

中間邊緣偵測法 ； 最低有效位元取代法 ； 最佳像素調整法

並列摘要

In this study, we propose an improved steganographic method based on median edge detection and optimal pixel adjustment process. At first, the median edge detection predictor distinguishes the target pixels located at the edge or those in the non-edge area, which determines the number of secret embedded data for each target pixel. According to the concept of human vision system, if the target pixel is located at the edge, the predicted value of the target pixel embeds more secret bits via the least-significant-bit substitution and optimal pixel adjustment process method. Contrarily, if the target pixel is located in the non-edge area, it embeds fewer secret bits via the same method. In this experiment, the modified method achieves a high capacity with better stego-image quality by using 8 grayscale images such as Lena, etc. This method does not require complex computational processes, but uses existing basic hardware and software to hide confidential information in the image and retrieve the hidden information from the stego-image. This study can provide reference applications in the field of defense information security management research.

並列關鍵字

Median Edge Detection ； Least-Significant-Bit Substitution ； Optimal Pixel Adjustment Process