Practical attacks on a class of secret image sharing schemes based on Chinese Remainder Theorem

Abstract

Chinese Remainder Theorem based secret image sharing schemes provide an efficient way to store secret image files securely in a distributed manner. Two basic properties required for designing such schemes are — availability and confidentiality. The first property ensures the retrieval of secret image from a qualified subset of shares and the latter guarantees the secrecy of the image against a coalition of non-qualified subset of shares. To model practical attacks on such schemes, in this paper, more powerful active adversaries are considered who can tamper with some shares to modify the reconstructed secret image. We show that an active adversary can tamper with one single share to make the honest parties reconstruct a false secret image of her choice. Our mathematical analysis along with the experimental data exhibit the practicality of our attacks for large class of existing schemes.