Toward a Secure and Robust Medical Image Watermarking in Untrusted Environment

Abstract

An untrusted environment such public Cloud as a model delivering hosted services through network. Several data are transmitted over the cloud which is considered as unsecured network, some of them may contain sensitive and important information such as medical images. In order to protect these images from several attacks, implementing a new robust digital image watermarking technique becomes necessary. In such techniques, a selected watermark is embedded in the medical image transmitted through network. The embedded watermark must be invisible and undetectable and only authorized persons can extract it to prove authentication. This paper propose a new hybrid image watermarking technique based on Number Theoretic Transform (NTT) and Diffie-Hellman on medical images in order to guarantee the security and preserve quality of medical images. We measured the robustness of the proposed approach by the commonly used metrics against several scenarios of attacks such as adding noise, rotation, etc. The tests are performed on several types of medical images.