作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (4): 7-9. doi: 10.3969/j.issn.1000-3428.2011.04.003

• 博士论文 • 上一篇    下一篇

基于八叉树空间分割的三维点云模型密写

綦 科,谢冬青,刘 洁   

  1. (广州大学计算机科学与教育软件学院,广州 510006)
  • 出版日期:2011-02-20 发布日期:2011-02-17
  • 作者简介:綦 科(1972-),男,副教授、博士研究生,主研方向:信息隐藏,图像处理,PKI/PMI;谢冬青,教授、博士、博士生导师;刘 洁,高级实验师、硕士
  • 基金资助:
    国家“863”计划基金资助项目(2009AA012420);广东省自然科学基金资助项目(9151009001000059)

3D Point Cloud Model Steganography Based on Octree Space Division

QI Ke, XIE Dong-qing, LIU Jie   

  1. (School of Computer Science and Education Software, Guangzhou University, Guangzhou 510006, China)
  • Online:2011-02-20 Published:2011-02-17

摘要: 针对三维点云模型的信息隐藏,提出一种基于八叉树空间分割的空域密写算法。对经过主成分分析后的三维点云模型建立包围盒,利用八叉树空间分割得到小体元并记录分割过程,通过顶点位移将信息嵌入到小体元内的不同空间位置。实验结果表明,该算法在提取信息时不需要原始模型数据,具有嵌入量高、失真度低的特点,能够抵抗旋转、平移、均匀缩放和顶点重排序攻击,适合于任意网格的三维模型信息隐藏。

关键词: 三维点云模型密写, 八叉树空间分割, 空域

Abstract: Aiming at the information hiding of 3D point cloud model, this paper presents a spatial domain steganography algorithm based on octree space division. Preprocessed by Primary Component Analysis(PCA), the bounding box of the 3D point cloud model is established and subdivided into voxel based on octree, in which octree structure is used to record segmentation process. The hiding messages are then embedded in the subspace of the voxel by moving the point position. Experimental results show that the extraction process of the algorithm does not need original data model when extracting information, it has features of high capacity and low distortion, it is robust against affine transformations and vertex reordering attacks and also suitable for any meshes of 3D model.

Key words: 3D point cloud model steganography, octree space division, spatial domain

中图分类号: