Research On Lossless Recovery Technologies Of Visual Secret Sharing

As smart phones,various handhold and the camera terminal equipment spread,the digital image acquisition,storage,and display presents an “explosion” status.And because of image data leakage and other security incidents are also common in recent years,people attach more importance to the safety of digital image data than ever before at present.The Visual Secret Sharing(VSS)technology,as more security and more effective technology which is used to protect media images,has been studied extensively.The traditional VSS recovers the secret images by stacking.Although the recovery phase is simple,only part recognizable information of original secret image is obtained.Obviously,the VSS can not apply to personal privacy image,military,and so on that need to be protected in detail.It is more imperious to study the lossless recovery technology for VSS.This dissertation focuses on lossless recovery technology for VSS.Aiming to solve the problem of high computational complexity and general threshold in the phase of lossless recovery of VSS,the lossless recovery algorithm based XOR operation and mapping are proposed.And based on which the lossless recovery scheme is proposed aim to the VSS with pixel expansion.Meanwhile,we study the homomorphism of the lossless recovery VSS.The main research works and innovations are summarized as follows:Firstly,a novel random grid-based(k,n)threshold VSS scheme is proposed.In the algorithm,the effective bits are used to increases the possibility to improve the contrast and the some bits are set to ‘0' to solve the question that the background of recovery image becomes dark with the shadow images increasing in some degree.Moreover,a general lossless recovery algorithm for VSS without pixel expansion is proposed.Theory analysis and simulation results show the effectiveness and security of the proposed scheme.And meanwhile,comparisons with previous related approaches show the advantages of the proposed algorithm and the scheme has more broad application prospectsSecondly,according to the proposed design concept of mapping theory,a novel mapping-based lossless recovery algorithm for VSS is proposed.The proposed scheme first implements the simple lossless recovery secret image on the condition of less than all n shadow images.And the definition of a new VSS named(k,tr,n)VSS is given.Through any subset of less k shadow images cannot get any information about the secret image,and the secret image can be revealed by any qualified subset of shadow images,i.e.,at least k shadow images.The secret image can be reconstructed losslessly by any tror more trshares.Thirdly,we propose a new lossless recovery algorithm based pixel merging for VSS schemes with pixel expansion.Most of the previous VSS schemes exist pixel expansion,and it is difficult to apply them to the practical application.According to the basic matrix,we can directly estimate whether the VSS has the lossless recovery ability or not.If the VSS scheme has lossless recovery ability,then the expanded secret image can be losslessly recovered by XOR or mapping operation.Moreover,the original secret image can be lossless recovered by the process of pixel merging which changes the multi pixels to one pixel.Fourthly,we study the homomorphism of the VSS with ability of lossless recovery.The extended attribute homomorphism of visual secret sharing is studied.The means that the homomorphism VSS is that the result of performing operation on the shadow images is the same as the result of performing operation on the secret image.In other words,the results obtained by superimposing or lossless recovery the operated shadows are exactly the same as those obtained by operating directly on the secret image.The homomorphism of the VSS can be applied to various practical scenes to solve the problem of secret image protection.The secret image is protected effectively when the generated shadow images are stored,transmitted and operated in the processing.