A robust privacy protection scheme for stereoscopic medical images based on fractal chaos

Taking three-dimensional (3D) two-wing chaotic system as the seed, we propose in this paper, a 3D four-wing chaotic system based on fractal. The dynamic characteristics of complex chaotic attractors are analyzed using the Lyapunov exponent and the Poincare cross-section. The results show that chaotic sequences with fractal transformation have high complexity and all NIST SP800-22 test data are greater than 0.01. A selective encryption scheme based on mask fusion for stereo medical images was proposed from the fractal chaos system. First, the stereo medical images were divided into upper, middle and lower parts of the image, from which, a portion of it was selected for the region of interest of the mask division. The three-part masks were merged and the entire stereo image was divided into background region and region of interest according to the mask. Then, the region of interest is extracted and converted into a new cube, and a cube rotation scrambling diffusion algorithm is designed. Finally, audio data containing the patient’s medical information and diagnostic information is embedded in the encrypted region of interest as a watermark. After testing and performance analysis, the algorithm has good encryption and watermarking performance, the histogram and scatter plot adheres to approximate uniform distribution, can resist various attacks, and has a higher security level and faster encryption speed.