A SURF and SVD-based robust zero-watermarking for medical image integrity

Area of Science:

• Medical Imaging
• Digital Security
• Image Processing

Background:

• Medical image security is critical in the digital age.
• Existing methods often compromise image quality or offer limited protection.
• There is a need for robust and non-intrusive medical image watermarking solutions.

Purpose of the Study:

• To introduce an innovative zero-watermarking methodology for medical images.
• To ensure robust protection of medical images without compromising diagnostic quality.
• To evaluate the methodology's performance against various attacks.

Main Methods:

• Utilized Sped-up Robust features (SURF) for high-precision feature extraction.
• Employed Singular Value Decomposition (SVD) for watermark embedding in the frequency domain.
• Developed a non-intrusive watermarking approach preserving original image integrity.

Main Results:

• Demonstrated superior performance and robustness against Gaussian noise, JPEG compression, and geometric distortions.
• Achieved high Normalized Correlation (NC) and Peak Signal-to-noise ratio (PSNR) values.
• Maintained near-perfect watermark resilience (NC close to 1.00) under Gaussian noise and rotation attacks, outperforming existing methods.

Conclusions:

• The proposed zero-watermarking methodology provides effective and imperceptible security for medical images.
• The technique preserves image diagnostic value while ensuring high resilience to various attacks.
• This approach represents a significant advancement in secure medical imaging practices.