Applying Chaotic System to Reversible Data Hiding

博士 === 國立彰化師範大學 === 工業教育與技術學系 === 106 === Using concepts from one of the most exciting areas of science, chaos theory, scientists can examine dynamic systems of uncertainty and create imaginative ways of addressing untracked behaviors. Chaos was first considered as disorder or confusion. The term of...

Full description

Bibliographic Details
Main Authors: Liu, Wei-Liang, 劉圍亮
Other Authors: Huang, Chuan-Kuei
Format: Others
Language:en_US
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/8v66cx
Description
Summary:博士 === 國立彰化師範大學 === 工業教育與技術學系 === 106 === Using concepts from one of the most exciting areas of science, chaos theory, scientists can examine dynamic systems of uncertainty and create imaginative ways of addressing untracked behaviors. Chaos was first considered as disorder or confusion. The term often induces misunderstandings on the nature of chaos theory. However, chaos generally refers to an approach or a methodology. Mathematicians and scientists study to control complex and complicated dynamic systems whose behavior are seemingly unpredictable. Data hiding is a technique that conceals data in a carrier for conveying secret messages. Digital images are transmitted widely over the Internet with high payload, and thus often serve as carriers. Secret data are embedded in the cover image, which is termed as stego-image with distortion. Reducing the image distortion is important in data hiding. Data hiding has two major requirements: imperceptibility and payload. Imperceptibility is estimated using the peak signal-to-noise ratio. High imperceptibility indicates low distortion between the stego and cover images. Payload is determined by the average amount of secret data embedded in each pixel of the cover image. In this dissertation, data hiding techniques for image files and their applications are investigated, and a novel method is proposed for grayscale image files. Given that adjacent pixels in natural images are strongly correlated, the author proposes a novel division mask instead of the traditional mask to exploit embeddable cases fully and increase data hiding capacities. A method is also proposed for hiding tremendous amounts of secret data in digital images. Natural images exhibit strong correlation between neighboring pixels, and thus the author proposes a block-based division mask instead of a cross division mask to exploit embeddable cases fully. The author also proposes a new encryption method based on logistic map and additive homomorphism to improve security. Experimental results show the proposed method achieves high payload and is immune to all kinds of attacks on the cryptosystem. Finally, a novel method is proposed for data hiding in encrypted images. Experimental results verify the feasibility and practicality of the proposed method. Keywords: block-based division; reversible data hiding; histogram shifting; logistic map