A Lossless Secret Sharing Scheme for Color Images

In this study, we introduce a new lossless visual secret sharing scheme (VSSS), inspired by Naor and Shamir’s pioneering visual cryptography scheme. While Naor and Shamir’s scheme uses Boolean algebra on binary images, this proposed method uses a transparent (acetate) base for the generation of each share and does not use Boolean algebra. Furthermore, the proposed method not only supports color images but also enables users to select the number of shares. For each share, selected pixels are randomly chosen from the secret image and placed at the exact location in a transparent base. To further reduce the visibility of the secret data shares, these shares are processed by utilizing the enhanced Arnold transform, thereby yielding the final noise-like share images via this hybrid algorithm. To evaluate the performance of the hybrid algorithm, image quality metrics were calculated. The results with the newly proposed scheme show that the secret image can be successfully reconstructed by merging the generated shares. The proposed method combined with the Arnold transform ensures that individual shares carry no intelligible information, while the combination of all shares allows for a perfect reconstruction of the original content. Experimental results and statistical analyses confirm the high correlation and homogeneity of the generated shares. Considering the existing studies in this field, the proposed hybrid VSSS contributes to the literature by providing a random pixel distribution and lossless image secret sharing. The proposed method successfully eliminates issues such as pixel expansion, contrast loss, and lossy data reconstruction using a newly proposed secret sharing scheme supported by Arnold’s Cat Map. The lossless nature of our algorithm is achieved as a direct result of the precise information-splitting process. The proposed method successfully addresses secure image splitting, storage, and subsequent lossless reconstruction.