A new color image secret sharing protocol

José Ignacio Farrán,David Cerezo
2023-06-21
Abstract:Visual cryptography aims to protect images against their possible illegitimate use. Thus, one can cipher, hash, or add watermarks for protecting copyright, among others. In this paper we provide a new solution to the problem of secret sharing for the case when the secret is an image. Our method combines the Shamir scheme for secret sharing using finite fields of characteristic 2 with the CBC mode of operation of a secure symmetric cryptographic scheme like AES, so that the security relies on that of the mentioned techniques. The resulting shares have the same resolution as that of the original image. The idea of the method could be generalized to other multimedia formats like audio or video, adapting the method to the corresponding encoded information.
Cryptography and Security,Information Theory
What problem does this paper attempt to address?
This paper proposes a new secret sharing protocol for color images, aiming to address how to securely divide an image (especially a color image) into multiple parts (called "shares") such that these shares individually contain no information about the original image. Only when a preset number of shares are combined can the original image be restored. This method falls under the category of visual cryptography. Specifically, the scheme combines Shamir's secret sharing scheme (based on polynomial interpolation over a finite field) and a secure symmetric encryption scheme (such as AES, using CBC mode). The purpose of this combination is to enhance security, ensuring that even if an attacker possesses some shares, they cannot infer any information about the original image. The characteristics of this method include: 1. **Shares of the same type as the original image**: The generated shares are also color images. 2. **Shares with the same resolution as the original image**: This makes each share look like a random image, without revealing any information about the original image. 3. **Shares are completely independent and random**: Each share appears random and does not contain any recognizable patterns. 4. **Different participants hold different shares**: Due to the use of different initialization vectors, each participant receives a unique share. The advantage of this scheme is that it is not only applicable to color images but can theoretically be extended to other multimedia formats (such as audio or video files). Additionally, the paper provides experimental results demonstrating the algorithm's efficiency with images of different sizes.