Résumé:
The rapid advancement of communication technologies creates significant chal-
lenges in protecting medical imaging data. Digital watermarking has emerged as a
formidable solution, embedding inconspicuous yet traceable information into mul-
timedia content to safeguard data security and authenticity, especially in health-
care. Unlike other methods, watermarking intrinsically links the information to
the data, providing robust protection against tampering and unauthorized modi-
fications.
In e-healthcare, the secure transfer of medical images is crucial for services such
as telemedicine and teleradiology, as it improves diagnostic accuracy and protects
patient data. This study introduces advanced blind watermarking methods based
on biometric technology to ensure secure transmission while maintaining image
integrity and authenticity. The proposed techniques are designed to balance the
practical challenges of detection rates, watermark visibility, and overall robustness.
This work presents robust watermarking frameworks to enhance medical image
management in e-healthcare systems, highlighting practical benefits and chal-
lenges. It integrates advanced security methods like encryption, blockchain, and
compression algorithms, which will be tested in realistic scenarios to validate their
effectiveness. By combining these innovative solutions, this research supports the
creation of secure and efficient systems for electronic medical records.
This thesis introduces two advanced watermarking frameworks founded on bio-
metric technology to ensure secure image transmission while preserving integrity.
The first, a Multi-Layered Security Framework (MLSF), integrates compressed
fingerprint watermarking with encryption and blockchain. The second approach,
Improved DWT and ACM Watermarking (IDAW), combines palmprint features
with chaotic transformations to enhance the watermark’s resilience.
Experimental evaluations confirm the effectiveness of these contributions. The
MLSF demonstrates exceptional imperceptibility, achieving a PSNR of 63.24 dB
and an SSIM of 1.0. The IDAW framework proves highly robust against a range of
attacks, including compression and noise, maintaining a PSNR of 53.95 dB and an
SSIM of 0.99996. By integrating innovative security solutions, this research con-
tributes to the development of secure and efficient systems for managing electronic
medical records in the modern digital era.
