Robust watermark technique using masking and Hermite transform.

Area of Science:

• Digital Image Processing
• Computer Vision
• Information Security

Background:

• Digital watermarking is crucial for copyright protection and data integrity.
• Existing methods often struggle with imperceptibility or robustness against attacks.
• Human visual system (HVS) properties can be leveraged for better watermarking.

Purpose of the Study:

• To evaluate a novel digital image watermarking algorithm.
• To ensure imperceptibility to the human eye and robustness against common processing and geometric attacks.
• To assess the algorithm's performance using various quantitative metrics.

Main Methods:

• A perceptive mask and normalization process were used for imperceptibility.
• The Hermite transform, based on Gaussian derivatives, was employed for image reconstruction.
• Blind watermark extraction was performed without the original image.
• Evaluation metrics included Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Normalized Cross-Correlation (NCC), and Bit Error Rate (BER).

Main Results:

• The algorithm demonstrated robustness against common processing and geometric attacks.
• Imperceptibility was achieved by preventing human eye detection.
• The study identified the maximum number of modifiable bits for adequate watermark extraction.
• Quantitative analysis confirmed the algorithm's effectiveness.

Conclusions:

• The proposed watermarking algorithm offers a good balance between imperceptibility and robustness.
• The Hermite transform effectively incorporates HVS properties for enhanced watermarking.
• The blind extraction capability makes the algorithm practical for real-world applications.