TY  - MANSCPT
AU  - Dhani Ariatmanto,
TI  - Robust and fast image watermarking scheme based on adaptive embedding strength
SN  - THE0009559 (Local)
PY  - 2021///
CY  - Kuantan, Pahang
PB  - UMP
KW  - Faculty of Computing
KW  - Dissertations
KW  - Universities and colleges
KW  - Theses
N1  - Faculty of Computing; Thesis (Doctor of Philosophy) -- Universiti Malaysia Pahang – 2021; Includes bibliographical references
N2  - Multimedia data such as images can be easily copied or manipulated by unauthorized users. Copyright protection is needed to avoid piracy, forgery and illegal distribution from unauthorized people. Digital watermarking has been widely used to protect intellectual properties. The current issues on the existing image watermarking include the scaling factor of the embedding watermark, imperceptibility, robustness, and time consumption during embedding of the watermark. The existing schemes apply the nonhybrid methods, hybrid methods and optimization methods to achieve an optimal embedding strength. However, the existing adaptive watermarking schemes do not achieve the satisfaction level in terms of imperceptibility, robustness, and computational time. Scaling factor plays an important role in balancing the robustness and imperceptibility of the watermarked image. Generating adaptive embedding strength based on its image content with less computational time becomes a solution to address this issue. In this approach, the image blocks that have the lowest variance of pixels have been chosen as the embedding regions. This study has proposed EWAES and FEWAES schemes to generate adaptive embedding strength by considering the impact of selected Discrete Cosine Transform (DCT) coefficients on its image block. The EWAES scheme examines the five selected DCT coefficients with a set of equations towards the average DCT to obtain adaptive embedding strength. The watermark is embedded by using a set of embedding rules. The experimental results show that the EWAES scheme produces high invisibility of the watermarked image with a peak signal-to-noise ratio (PSNR) value of 46.69 dB and a Normalized Cross-correlation (NC) value of 0.973. The FEWAES scheme is designed to achieve a fast-embedding watermark by considering the selected two DCT coefficients towards the average DCT coefficients. The FEWAES scheme achieved the least computational time with average time consumption of 0.141 seconds for embedding a watermark. The proposed FEWAES scheme has the potential to be implemented in mobile devices
ER  -