Scholarlink

<p>As information technology has quickly developed, many different industries have emerged along with it. Fields using 3D models as their fundamental data formats are no exception. This study examines applying reversible data hiding techniques of encrypted 2D media to 3D mesh models for privacy protection purposes. Examining the unique characteristics of 3D models and how they differ from 2D media can offer us ways to innovate in more diverse directions using some state-of-the-art algorithms as the foundation to improve the privacy of the original mesh models. The proposed algorithm uses a circulated method to gradually diffuse the reference data set which is essential to be able to recover the original mesh model. It involves coordinate transformation, randomly and recursively selecting reference sets, and determining the embeddable capacities according to prediction error detection process on multi-MSBs. Furthermore, our proposed scheme compares the results of arithmetic coding and Huffman coding which are used to compress auxiliary data. Finally, auxiliary information used for message extraction and model recovery is embedded by simple bit substitution. Experimental results demonstrate that the proposed method achieves an average embedding rate of 49.08 bits per vertex (bpv), representing a significant improvement over the compared schemes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Reversible data hiding in encrypted 3D mesh models via reference vertex circulation strategy

  • 期刊论文

  • 发表日期：

    2025年5月31日

查看全文
  • Jui-Chuan Liu,
  • Ching-Chun Chang,
  • Kai Gao,
  • Chin-Chen Chang

摘要

As information technology has quickly developed, many different industries have emerged along with it. Fields using 3D models as their fundamental data formats are no exception. This study examines applying reversible data hiding techniques of encrypted 2D media to 3D mesh models for privacy protection purposes. Examining the unique characteristics of 3D models and how they differ from 2D media can offer us ways to innovate in more diverse directions using some state-of-the-art algorithms as the foundation to improve the privacy of the original mesh models. The proposed algorithm uses a circulated method to gradually diffuse the reference data set which is essential to be able to recover the original mesh model. It involves coordinate transformation, randomly and recursively selecting reference sets, and determining the embeddable capacities according to prediction error detection process on multi-MSBs. Furthermore, our proposed scheme compares the results of arithmetic coding and Huffman coding which are used to compress auxiliary data. Finally, auxiliary information used for message extraction and model recovery is embedded by simple bit substitution. Experimental results demonstrate that the proposed method achieves an average embedding rate of 49.08 bits per vertex (bpv), representing a significant improvement over the compared schemes.