Zero-Watermarking Algorithm for Medical Images Based on Nonsubsampled Contourlet Transform and Double Singular Value Decomposition

The privacy and security of medical images in storage and transmission are enormous challenges. In this paper, we propose a novel multi-algorithm fusion of medical image robust zero-watermarking algorithm based on the nonsubsampled contourlet transform (NSCT), double singular value decomposition (DSVD), and multi-level discrete cosine transform (MDCT). The NSCT of the original medical image obtains low-frequency domain information, blocks the low-frequency domain information and uses the MDCT to obtain the coefficient matrix of the low-frequency direction sub-bands, and then uses the DSVD to construct the feature vector. At the same time, the watermarking is encrypted using Logistic Map to ensure the security of the original watermarking information under the chaotic system. In the watermarking embedding and extraction phase, zero-watermarking technology is used to ensure the integrity of medical images. Experimental results show that the proposed algorithm can extract watermarking information effectively and accurately with fast calculation speed and high precision, and it has good invisibility and strong robustness for both common attacks and geometric attacks.