Research On Optimal Of Visual Cryptography

Visual cryptography is a new cryptography branch, which is gradually developed in recentten years. It is a combination of images and one-time pad that is a traditional encryptionmethod. The decryption can be implemented by direct using human vision. Therefore, it is anabsorbing research. The main idea of visual cryptography is using the idea of secret sharing.A black-and-white image can be shared to n chaotic and meaningless images that are calledshares or shadows. Afterwards, the n shares are distributed to n participators. In thedecryption process, k (k n)shares should be collected and superimposed and then the secretimages can be directly shown by human vision. However, the reconstructed images aredistortion and expansion. Therefore, the subsequent studies mainly concentrate on the schemeoptimization of visual cryptography and the applications of visual cryptography.This paper designs two practical visual cryptography schemes based on the optimization ofvisual cryptography. One is a multi-secret images scheme based on Boolean Operation, whichis used to share n secret images with different dimensions to n share images with the samedimension. The reconstruction is simple with easy computations. Moreover, the proposedscheme can deal with grayscale, color, and the mixed condition of grayscale and color images.Furthermore, the proposed scheme can be used to handle multi-images with differentdimensions. The reconstructed images without expansion and the image contrasts are one.Another scheme is XOR-based visual cryptography for grayscale images, which is used toencrypt the grayscale secret images. In the proposed scheme, a secret image can be shared ton share images. In the sharing process, the encryption items can be chosen according to therandom numbers that are generated by a random number generation algorithm. Therefore,each encryption item is random, similar with those of one-time pad schemes.In this paper, C program language is used to simulation the proposed two schemes. Theresults have strong practical significance. The two schemes proposed in this paper arecompared with other schemes in pixel expansion, contrast, secret images’ dimensions andrandomize, respectively. The results show that schemes in this paper not only optimize theprevious schemes, but also have practical significance.