Multilevel Security Framework with Fault Tolerance Using Secret Image Sharing and Steganography

Secure data embedding within digital images has become crucial for protecting sensitive information against unau-thorized access. In this work, we propose a multilevel security that integrates data embedding, two-level secret sharing, and message extraction within digital images. The embedding process involves encoding data into grayscale images using random least significant bit substitution, ensuring covert integration while preserving image quality. Subsequently, a multilevel secret sharing scheme based on Shamir's secret sharing is applied to generate multiple shares of the image, enhancing security through distributed storage and threshold-based reconstruction. The extraction phase employs polynomial interpolation to reconstruct the original image from the shares, facilitating seamless retrieval of embedded data. The adversary will not be able to obtain any information about the secret without significant computation cost.