Find in Library

Digital watermarking can be applied to protection of medical images privacy, hiding of patient's diagnosis information and so on. In order to improve the ability of resisting geometric attacks, a new watermarking algorithm for medical volume data in sub-block three-dimensional discrete cosine transform domain is presented. The original watermarking image is scrambled by a Chebyshev chaotic neural network so as to improve watermarking security. Sub-block three-dimensional discrete cosine transform and perceptual hashing are used to construct zero-watermarking. In this way it does not produce medical image distortion and gives the algorithm the ability to resist geometric attacks. Experimental results show that the algorithm has good security, and it has good robustness to various geometric attacks.