Consistency Analysis for Secure Reconstruction in Secret Image Sharing
Anneke Soraya Hidayat, Gil‐Je Lee, Eun‐Jun Yoon, Kee-Young Yoo · 2015
Secret sharing is a scheme to distribute a secret message s into n participants, thus t or more participants can reconstruct the secret message s. While less than t shares can know nothing about the secret, the problem arises when there are more than t shares. The adversary A can use this to put their share in the last. This paper presents the analysis of consistency in secret image sharing when there are m participants, where t ≤ m ≤ n and adversaries A among the reconstruction subset. The dealer collects t different pixels from the subset t out of the set of m repetitively. The results show that the higher threshold t can produce lower image quality when the adversary A is among of the subset. Hence, the dealer D can visually detect the adversary A when m ≤ 2t -- 1. It shows that the necessity of consistency is sufficient to determine the adversary A.