Research On Multi-secret Sharing Visual Cryptography Scheme

As a new image sharing technology, Visual Cryptography (VC) can recover secret images using human visual system. Meanwhile, the"one time pad"security makes researchers focus on VC in recent years. Multi-secret sharing visual cryptography scheme (MVCS) is an important part of VC. MVCS not only increases the count of shared secret images, but also can help to modify the VC theory and extend the application domain. This thesis reviews the research advances of VC, and analyzes the main kinds of MVCS: Access-based MVCS (AMVCS) and Operation-based MVCS (OMVCS). Since their shortcomings, some advanced schemes are designed in this thesis. A new MVCS is defined and the corresponding scheme is designed, in order to adapt the common relationships between qualified sets and secret images.Firstly, this thesis points out that the authorities of participant sets are obscure in AMVCS with threshold structure. Based on the ( k , l , n ) ? VCS, which denotes the single secret VCS with upper and lower thresholds, a AMVCS is defined with clear qualified sets of all secret images. The construction algorithm of responding basic matrices are also designed. Compared with the previous ones, the pixel expansion of this scheme is smaller and the visual effects are better. The large pixel expansion and small relative difference are common problems of OMVCS.This thesis proposes an advanced scheme for ring shares. Based on the relative matrices, the secret sharing and recovering algorithms for ring shares are designed. The proposed scheme can share WT (the width of secret images) images at most. Furthermore, the pixel expansion reduces from 3h (h is the number of secret images) to 2h and the relative difference increases from 1 (3h ) to 1 (2 h ), which obviously improve the visual effects of recovery images.It is discovered in this thesis that the general construction based on Mask causes the unequal statuses of two shares in OMVCS. It is the key point of the limited participant number. A new OMVCS without Mask is designed using single secret VCS and the shares'marked areas. The scheme realizes the equal statues of two shares, and holds the advantages of secret image number and recovery effects. The idea provides a new way for increasing the participant number, and also makes it possible to unify AMVCS and OMVCS.Although AMVCS and OMVCS can share multiple secret images with different principles, the application scopes of two schemes are both limited. Finally, this thesis defines a new MVCS, which can adapt any relationship between qualified sets and secret images. AMVCS and OMVCS are special statuses of the new definition. Combining the basic matrices and areas marking, a scheme corresponding to the definition is designed. The effectiveness of the scheme is proved by both proof and experiment. The contents of this thesis are beneficial to advancing MVCS and laying the foundation for the applications of VC.