An adaptive robust watermarking scheme based on chaotic mapping.

Area of Science:

• Information Security
• Digital Image Processing
• Cryptography

Background:

• Digital image transmission faces increasing security risks.
• Image watermarking is crucial for protecting image information security.
• Balancing imperceptibility, robustness, and embedding capacity is a key challenge in image watermarking.

Purpose of the Study:

• To propose a high-capacity color image adaptive watermarking scheme.
• To enhance the security, imperceptibility, and robustness of digital image watermarking.
• To improve embedding capacity and reduce processing time.

Main Methods:

• Utilized Discrete Wavelet Transform (DWT), Heisenberg Decomposition (HD), and Singular Value Decomposition (SVD).
• Employed Logistic chaotic mapping for watermark encryption.
• Adaptive embedding factor calculation via cover image entropy and Alpha blending in YCbCr color space.
• Incorporated singular value correction for enhanced robustness.

Main Results:

• Achieved high imperceptibility with average PSNR of 45.3437dB and SSIM of 0.9987.
• Demonstrated strong robustness against various attacks, with average NCC above 0.95.
• Reached a high embedding capacity of 0.6667bpp and an average running time of 1.1136 seconds.

Conclusions:

• The proposed watermarking scheme effectively balances imperceptibility, robustness, and high embedding capacity.
• Offers superior performance compared to existing watermarking techniques in terms of capacity and speed.
• Provides a secure and efficient solution for digital image information protection.