Application of Virtual Reality Technology and Unsupervised Video Object Segmentation Algorithm in 3D Model Modeling

IF 1 4区物理与天体物理 Q3 PHYSICS, MATHEMATICAL

Advances in Mathematical Physics Pub Date : 2022-10-13 DOI:10.1155/2022/4743456

Hui-Lin Yang, Qiuming Liu

{"title":"Application of Virtual Reality Technology and Unsupervised Video Object Segmentation Algorithm in 3D Model Modeling","authors":"Hui-Lin Yang, Qiuming Liu","doi":"10.1155/2022/4743456","DOIUrl":null,"url":null,"abstract":"3D modeling is the most basic technology to realize VR (virtual reality). VOS (video object segmentation) is a pixel-level task, which aims to segment the moving objects in each frame of the video. Combining theory with practice, this paper studies the process of 3D virtual scene construction, and on this basis, researches the optimization methods of 3D modeling. In this paper, an unsupervised VOS algorithm is proposed, which initializes the target by combining the moving edge of the target image and the appearance edge of the target and assists the modeling of the VR 3D model, which has reference significance for the future construction of large-scale VR scenes. The results show that the segmentation accuracy of this algorithm can reach more than 94%, which is about 9% higher than that of the FASTSEG method. 3D modeling technology is the foundation of 3D virtual scene; so, it is of practical significance to study the application of 3D modeling technology. At the same time, it is of positive significance to use the unsupervised VOS algorithm to assist the VR 3D model modeling.","PeriodicalId":49111,"journal":{"name":"Advances in Mathematical Physics","volume":" ","pages":""},"PeriodicalIF":1.0000,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Advances in Mathematical Physics","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1155/2022/4743456","RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"PHYSICS, MATHEMATICAL","Score":null,"Total":0}

引用次数: 0

Abstract

3D modeling is the most basic technology to realize VR (virtual reality). VOS (video object segmentation) is a pixel-level task, which aims to segment the moving objects in each frame of the video. Combining theory with practice, this paper studies the process of 3D virtual scene construction, and on this basis, researches the optimization methods of 3D modeling. In this paper, an unsupervised VOS algorithm is proposed, which initializes the target by combining the moving edge of the target image and the appearance edge of the target and assists the modeling of the VR 3D model, which has reference significance for the future construction of large-scale VR scenes. The results show that the segmentation accuracy of this algorithm can reach more than 94%, which is about 9% higher than that of the FASTSEG method. 3D modeling technology is the foundation of 3D virtual scene; so, it is of practical significance to study the application of 3D modeling technology. At the same time, it is of positive significance to use the unsupervised VOS algorithm to assist the VR 3D model modeling.

查看原文本刊更多论文

虚拟现实技术和无监督视频对象分割算法在三维模型建模中的应用

3D建模是实现VR(虚拟现实)的最基本技术。视频对象分割(VOS)是一种像素级的任务，其目的是分割视频中每一帧的运动对象。本文结合理论与实践，研究了三维虚拟场景的构建过程，并在此基础上研究了三维建模的优化方法。本文提出了一种无监督VOS算法，通过结合目标图像的运动边缘和目标的外观边缘对目标进行初始化，辅助VR三维模型的建模，对未来大规模VR场景的构建具有参考意义。结果表明，该算法的分割准确率可达到94%以上，比FASTSEG方法提高约9%。三维建模技术是三维虚拟场景的基础;因此，研究三维建模技术的应用具有重要的现实意义。同时，利用无监督VOS算法辅助VR三维模型建模也具有积极意义。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Advances in Mathematical Physics 数学-应用数学

CiteScore

2.40

自引率

8.30%

发文量

151

审稿时长

>12 weeks

期刊介绍： Advances in Mathematical Physics publishes papers that seek to understand mathematical basis of physical phenomena, and solve problems in physics via mathematical approaches. The journal welcomes submissions from mathematical physicists, theoretical physicists, and mathematicians alike. As well as original research, Advances in Mathematical Physics also publishes focused review articles that examine the state of the art, identify emerging trends, and suggest future directions for developing fields.