In this paper, the effects of noise attacks on the performance of a hyperchaos-based digital image encryption algorithm is evaluated. The effect of noise degradation affects the visual information fidelity, structural content and the decryption performance of an image encryption algorithm. Images are susceptible to degradation by noise from different sources as they undergo various processes. Robustness analyses have also been performed on several image algorithms including those based on Josephus traversing and mixed chaotic map using salt and pepper noise, partial image encryption using gaussian noise and salt and pepper noise, image encryption based on intertwining logistic map using Gaussian noise, image encryption based on DNA complementary rules using gaussian noise and salt and pepper noise and image encryption scheme based on stream cipher using a random noise, amongst others. the decrypted image indicated that the algorithm is robust as the decrypted image was identical to the encrypted version. In the work in, the authors evaluated the robustness of an image encryption algorithm based on Kenken puzzle, in which a Lena image was degraded with salt and pepper noise and speckle noise respectively with variance of 0.01. Several types of noises such as blur, Gaussian white, salt and pepper, speckle, Poisson, motion blur, erosion and dilation noises can be added to test images, prior or during encryption operations to evaluate robustness. Besides, the simulation and the performance analyses with varied validation metrics express the robustness, defiance to the multifarious threats and potential for some real world application of the cipher. The value of information entropy attained through the proposed algorithm is 7.9975 which is very competitive. #Pixel puzzle generator updateIn order to introduce the plaintext sensitivity in the proposed cipher, the hash codes given by the SHA-256 hash function for each input image have been used to update the system parameters and the initial values of the chaotic systems. Our study swapped randomly selected portions of the rows and columns for scrambling, which spawned more security. In the previous studies, whole rows and whole columns were swapped with each other. An Exclusive-OR(XOR) operation is conducted between the scrambled image and the keystream given by PWLCM to get the final encrypted image. Resultantly, the given image is confused abundantly. If selected vectors of pixels intersect each other, then to avoid the loss of pixels data, an intersection filter (IF) has been employed to forgo the intersecting vectors of selected pixels from the row and column for swapping. A randomly selected vector of pixels from these selected rows and columns is swapped with each other for a number of times. After the grayscale image is input, arbitrary row and column are selected from this image to realize the scrambling effects. Two more streams of random numbers have been created by manipulating the two streams of Henon map. These maps provide three streams of random numbers. Two chaotic maps, Henon map and piecewise linear chaotic map (PWLCM) have been used. Security of the cipher would be improved in this way. To introduce more complications for the potential hackers, a novel grayscale image encryption scheme using the variable length row-column swapping based scrambling (VLRCSBS) has been proposed in this study. Swapping a row with a row, a column with a column or row with a column are usual operations carried out to scramble the given input plain image.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |