A Hyper-Chaotically Encrypted Robust Digital Image Watermarking Method with Large Capacity Using Compress Sensing on...

Area of Science:

• Computer Science
• Information Security
• Digital Image Processing

Background:

• Existing digital watermarking techniques often struggle to balance robustness and capacity simultaneously.
• Image copyright protection and secure transmission remain critical challenges in digital multimedia.

Purpose of the Study:

• To propose a robust semi-blind image watermarking scheme with high capacity and improved security.
• To address the limitations of current methods in achieving both robustness and high data embedding capacity.

Main Methods:

• Discrete Wavelet Transform (DWT) for carrier image decomposition.
• Compressive sampling for watermark image compression.
• Combination of One and Two-Dimensional Chaotic Map (TL-COTDCM) for secure scrambling.
• Singular Value Decomposition (SVD) for watermark embedding.

Main Results:

• Successfully embedded eight 256x256 grayscale watermark images into a 512x512 carrier image, achieving 8x higher capacity than average existing methods.
• Demonstrated high robustness against common attacks, validated by Normalized Correlation Coefficient (NCC) and Peak Signal-to-Noise Ratio (PSNR) metrics.
• Achieved superior performance in robustness, security, and capacity compared to state-of-the-art watermarking techniques.

Conclusions:

• The proposed digital watermarking scheme offers significant improvements in capacity, robustness, and security.
• The method shows great potential for future multimedia applications requiring secure data embedding.
• The integration of DWT, compressive sampling, chaotic maps, and SVD provides a powerful approach to digital watermarking.