Visually meaningful image encryption scheme based on new-designed chaotic map and random scrambling diffusion strategy

Since most of the existing image encryption algorithms hide the plaintext information by encrypting the image into a noise like image, the encrypted image is very different from the original image, and the visual security of the encrypted image is ignored. In view of this defect, a visually meaningful image encryption scheme combining new chaotic mapping and parallel compressed sensing is proposed, which makes the image content and vision get double protection. Firstly, a newly designed one-dimensional chaotic map derived from the fractional form of trigonometric function is used for constructing the controlled measurement matrix and disturb the strong correlation of images. Secondly, a bidirectional random scrambling algorithm based on chaotic magic transformation (CMT) is designed to disturb the wavelet packet coefficient matrix of plain image. Then, the scrambled secret image is encrypted and compressed with the help of chaotic measurement matrix and the newly designed random chaotic pixel diffusion strategy. Finally, the secret image is hidden in another large carrier image such as plain image through the wavelet packet transform embedding algorithm. As a result, an image is protected by both visual and information security. Simulation results and security analysis make clear that the proposed encryption scheme is feasible and effective under the premise of guaranteeing visual security and decryption quality.