Master'sOpen Access

Steganografi kullanarak görüntü güvenlik yöntemi

2018
0 views
0 downloads
Advisor: Doç. Dr. Sema Kayhan

Abstract (EN)

One of the most important factors of information technology and communication is the security of the information. This thesis presents and compares two main steganography techniques; quad-tree based steganography and peak-valley methods. The quad-tree algorithm can be considered as an improved version of the traditional histogram shifting method. The peak-valley algorithm uses the advantage of simple local distribution of pixel intensities and utilizes the zero or the minimum points of the histogram of an image. The method slightly modifies the pixel grayscale values to embed the data into the image. It can embed more data than many of the existing reversible data hiding algorithms. The performance of the presented methods, in terms of imperceptibility, the embedding rate ranges from 200 to 1000 bytes, whenever six variants images are employed as cover images. The results have revealed that the imperceptibility for both algorithms is almost identical whereby, the PSNR decreases as the embedding rate increases. The PSNR was 44.2092 dB when the embedding rate was 200 bytes and gradually decreased as embedding rates increased. Furthermore, the PSNR of Peak-valley is higher than that of Quad-tree with the embedding rate ranges from 200 to 1000 .In addition, Chi-Square attack is performed on the stego image, produced by the presented algorithms to measure the robustness of the methods. The test reveals that the probability of containing secret message is almost zero for the entire test which indicates that, there is no hidden message inside the cover image. Key words: Embedding capacity maximization, Histogram, PSNR, Reversible data hiding

Author

Awab Qasım Karamanjı

How to Cite

Awab Qasım Karamanjı (Master Thesis). Steganografi kullanarak görüntü güvenlik yöntemi, 2018, Gaziantep University.

Keywords

License

Tüm Hakları Saklıdır

This work is shared under the specified license terms.

More theses from Gaziantep University