基于点云与NeRF的三维重建系统  

Three-Dimensional Reconstruction System Based on Point Cloud and NeRF

作　　者：曹晨曦 任泰安[2] CAO Chenxi;REN Taian(School of Computer Science and Information Engineering,Hefei University of Technology,Hefei Anhui 230601;Engineering Quality Education Center of Undergraduate College Hefei University of Technology,Hefei Anhui 230601)

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230601 [2]合肥工业大学本科生院工程素质教育中心,安徽合肥230601

出　　处：《软件》2024年第4期11-14,共4页Software

基　　金：合肥工业大学创新创业教育精品课程(14020-03322022001)。

摘　　要：三维重建可以帮助人们更好地查看被重建对象的细节,广泛应用于物体展示、虚拟场景模拟、文物保护等场景中。传统的多视图三维重建容易出现大量噪声、空洞等问题,而且展示效果不佳。新兴的利用神经网络进行三维重建的NeRF算法虽然能够达到接近真实的效果,但是渲染的速度慢,无法满足交互的要求。文章通过结合两种方法,得到了一种兼顾交互效果与渲染效果的三维重建系统。结果表明,本文系统在采集的稀疏图片数据集上,达到了较好的交互与渲染效果。3D reconstruction can help people better view the details of the reconstructed object and is widely used in object display,virtual scene simulation,cultural relic protection and other scenes.Traditional multi view 3D reconstruction is prone to a large amount of noise,voids,and other issues,and the display effect is poor.The emerging NeRF algorithm that utilizes neural networks for 3D reconstruction can achieve near real results,but the rendering speed is slow and cannot meet the requirements of interaction.The article combines two methods to obtain a 3D reconstruction system that balances interaction and rendering effects.The results show that the system in this article achieved good interaction and rendering effects on the sparse image dataset collected..

关 键 词：三维重建 NeRF 点云 WEB技术 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]