首页|针对顶点策略优化的三维模型可逆数据隐藏算法

针对顶点策略优化的三维模型可逆数据隐藏算法

扫码查看
随着3D打印技术和云计算的普及发展,三维模型的数据安全和隐私安全日益受到重视,研究模型无损恢复和存储海量数据的隐藏方法具有重要意义。基于此,以三维网格模型为载体,提出一种顶点策略优化的可逆数据隐藏方法。首先,对网格模型顶点进行划分,在全邻域中生成新的重心集合,根据改进的预测策略预测所有顶点集合的MSB以确定最大嵌入位;其次,在接受端应用顶点处拉普拉斯算子得到光滑恢复模型;最后,采用平均曲率可视化分析。为证明该算法的有效性,选取不同大小的网格模型与其他传统算法的嵌入率、SNR以及恢复模型进行对比。结果表明,该算法不仅提高了嵌入率,还在保留恢复模型局部细节特征的同时有效去除噪声,提高了模型恢复质量和视觉效果,进一步达到数据隐私安全的目的。
Reversible Data Hiding Algorithm for 3D Model Optimized for Vertex Strategies
With the popular development of 3D printing technology and cloud computing,data security and privacy of 3D mesh models are increasingly important,and it is important to study the methods of model lossless recovery and storage of massive data hiding.Based on this,we propose a reversible data hiding method with vertex strategies optimization using 3D mesh model as the carrier.Firstly,we divide the vertices of the grid model,generate a new set of centers of gravity in the full neighborhood,and predict the MSB of all the vertex sets according to an improved prediction strategy to determine the maximum embedding position.Then,apply the Laplace operator at the vertices to obtain a smooth recovery model at the receiver side.Finally,visualize and analyze the mean curvature.To prove the ef-fectiveness of the proposed algorithm,the embedding rate,SNR and recovery model of different sizes of lattice models are compared with those of other traditional algorithms.The results show that the proposed algorithm not only improves the embedding rate,but also effectively removes the noise while preserving the local detail features of the recovered model,which improves the quality and visual effect of the recovered model and further achieves the purpose of data privacy and security.

reversible data hiding3d modelvertex strategiesLaplace operatormean curvaturedenoising smooth

张国有、米佳

展开 >

太原科技大学 计算机科学与技术学院,山西 太原 030024

可逆数据隐藏 三维模型 顶点策略 拉普拉斯算子 平均曲率 去噪光滑

山西省自然科学基金资助项目山西省研究生教育创新项目

201801D1211332021Y699

2024

计算机技术与发展
陕西省计算机学会

计算机技术与发展

CSTPCD
影响因子:0.621
ISSN:1673-629X
年,卷(期):2024.34(1)
  • 30