A hybrid approach combining data hiding with visual cryptography for secure extraction of data hiding

Visual cryptography scheme (VCS) naturally provides a solution for protecting a secret image in a group environment. However, most of the existing VCSs proposed so far only protect one secret image. In this work, we design a hybrid scheme combining VCS and data hiding to simultaneously safeguard the secret image and some related information (e.g., the description of the secret image). First of all, a framework for the hybrid scheme is introduced. Algorithms for building the proposed framework are given. Generally, a secret image is partitioned into k shadows in which the additional messages are embedded. By stacking k shadows, the secret image is visually revealed, and meanwhile, the embedded messages can be securely extracted from the stacked result. Moreover, the performance of the proposed technique would suffer as a result when the threshold k becomes larger. At this time, we have larger pixel expansion, smaller contrast, and reduced data hiding efficiency. To make the scheme more practical, an improved approach is presented for reducing the pixel expansion, increasing the visual quality of recovered image, and enhancing the data hiding efficiency. Comprehensive theoretical analysis and experiments are demonstrated to clarify the effectiveness and advantages of the proposed methods.