| As the era of big data and the era of 5G come,the number and types of digital media on the network are growing rapidly,and the corresponding security problems are also emerging one after another.Attackers can tamper with,copy and destroy digital media without users’ knowledge,which makes the copyright issue more prominent.With the more extensive application of 3D model,the information security problem in 3D domain has become more and more prominent,and the information hiding technology based on 3D model has become an effective way to ensure the information security in 3D domain.3D model has the characteristics of large amount of data and rich types of information,and 3D point cloud model has the advantages of displacement invariance and rotation invariance.Based on the three-dimensional point cloud model,this paper deeply studies the preprocessing method,the preprocessing design rules of secret information,the optimal matching between information and the carrier and the information embedding rules,and puts forward two information hiding algorithms based on the point cloud model:(1)Information hiding algorithm based on point cloud increment.Firstly,the three-dimensional model is segmented by plane fitting clustering.Secondly,the point clouds of each layer are compressed by removing points,and the points of the point clouds are classified according to the clustering results,and the points with strong robustness are the information hiding areas;Finally,using the spatial characteristics between different point sets,the double embedding of secret information is completed by vertex segmentation.(2)A multi-carrier information hiding algorithm based on point cloud convex hull fusion is proposed.Firstly,a convex hull is established for multiple point cloud model carriers,and the multi-carrier fusion state is generated according to the center of the convex hull.Secondly,based on Morse theory,the points in the fusion state point cloud are divided to get the embedding area of secret information.Finally,the double matching modification of secret information and point cloud information is completed by modifying the intensity value of point RDB channel.The information hiding algorithm based on point cloud increment in this paper is simulated and compared with the experiment.The invisibility Hausdorff distance of the algorithm is reduced by 5.59% and 24.18% on average,and the BCR value of the algorithm is increased by 4.3% and 14.88% compared with the comparison algorithm in the face of point cloud simplification,uneven compression,random noise addition and point cloud shearing attacks.4.72%、15.39%;3.97%,15.16% and 3.37%,12.26%.Compared with the contrast algorithm,the multi-carrier information hiding algorithm based on point cloud convex hull fusion has optimized the geometric indexes of invisibility,such as MRSM and MSDM,and the color index P-PSNR tends to be stable with the increase of embedded information.In the aspect of robustness,the BCR value of the algorithm is higher than that of the contrast algorithm in the face of uneven scaling attack,random noise attack,model simplification attack and model shearing attack.10.41%、8.16%;6.21%、8.37%;6.67% and 7.53% and9.96% and 10.81%.To sum up,in this paper,the two point cloud information hiding algorithms have broken the balance between invisibility,robustness and capacity in both single-carrier domain and multi-carrier domain to some extent,and have certain application value in the application scenarios of single-carrier and multi-carrier high-stick,which has a positive impact on national network security. |
