Selection of the appropriate cover to hide information in images

Other Title(s)

اختيار الغطاء الملائم لإخفاء معلومات في الصور

Dissertant

Farahat, Rima F.

Thesis advisor

al-Mashaqibah, Firas Faris Musa

Comitee Members

al-Hamami, Ala Husayn
Fadil, Muayyad Abd al-Razzaq

University

Amman Arab University

Faculty

Collage of Computer Sciences and Informatics

Department

Department of Computer Science

University Country

Jordan

Degree

Master

Degree Date

2013

English Abstract

Steganography is the art and science of writing hidden messages in such a way that no one, apart from the sender and intended recipient, suspects the existence of the message.

In steganography the freedom to choose a cover that result in the least detectable stego image is an advantage that can be maximized.

In this thesis, we proposed a new cover selection approach with four versions then we compared between them based on their ability to minimize the detectability of the resulting stego image while keeping the host image distortion as minimum as possible.

Our cover selection technique versions are created by dividing the cover images and the secret image into blocks of size 4x4 and each block is divided into sub-blocks of different sizes (1x4, 4x1, 2x4, and 4x2); for each block we calculated the block texture and neighborhood information to measure the similarity of 256 blocks for the secret image as a one unit with the 4096 blocks for each cover image divided in 16 groups.

Finally; the best similar cover for each version is chosen to hide the secret image using the least significant bit (LSB) technique.

Our proposed methods were examined with wavelet based domain (WBS) and Fisher Linear Discriminator (FLD) to prove their robustness against detectability of the resulting stego image, also we examined the Peak signal-to-noise ratio(PSNR) to prove the ability of the four versions in maintaining minimum distortion in the host image.

The results show that (1×4) blocking gives the best security and the best minimum distortion, (2×4) was the fastest version, and (1×4, 4×1, 2×4 and 4×2) were simple.

Experimental results are carried out on image processing ToolBox in Matlab under windows 7, and images are randomly selected from Washington University image database by using method of the class random sample.

Main Subjects

Mathematics

Topics

• Encryption
• Digital electronics
• Data

No. of Pages

57

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-525930