MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo

This study addresses the challenge of online 3D model generation for neural rendering using an RGB image stream. Previous research has tackled this issue by incorporating Neural Radiance Fields (NeRF) or 3D Gaussian Splatting (3DGS) as scene representations within dense SLAM methods. However, most s...

Full description

Saved in:
Bibliographic Details
Main Authors: Byeonggwon Lee, Junkyu Park, Khang Truong Giang, Sungho Jo, Soohwan Song
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11050405/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849427927950688256
author Byeonggwon Lee
Junkyu Park
Khang Truong Giang
Sungho Jo
Soohwan Song
author_facet Byeonggwon Lee
Junkyu Park
Khang Truong Giang
Sungho Jo
Soohwan Song
author_sort Byeonggwon Lee
collection DOAJ
description This study addresses the challenge of online 3D model generation for neural rendering using an RGB image stream. Previous research has tackled this issue by incorporating Neural Radiance Fields (NeRF) or 3D Gaussian Splatting (3DGS) as scene representations within dense SLAM methods. However, most studies focus primarily on estimating coarse 3D scenes rather than achieving detailed reconstructions. Moreover, depth estimation based solely on images is often ambiguous, resulting in low-quality 3D models that lead to inaccurate renderings. To overcome these limitations, we propose a novel framework for high-quality 3DGS modeling that leverages an online multi-view stereo (MVS) approach. Our method estimates MVS depth using sequential frames from a local time window and applies comprehensive depth refinement techniques to filter out outliers. The refinement method produces temporally consistent depths by checking sequential geometric consistency, enabling accurate initialization of Gaussians in 3DGS. Furthermore, we introduce a parallelized backend module that optimizes the 3DGS model efficiently, ensuring timely updates with each new keyframe. Experimental results demonstrate that our method outperforms state-of-the-art dense SLAM methods, achieving an average PSNR improvement of approximately 2 dB on indoor scenes. Moreover, our method reliably produces consistent 3D models in complex outdoor scenes, where existing methods often fail due to tracking errors and depth noise. It also reconstructs large-scale aerial scenes effectively, achieving an average PSNR gain of about 10.28 dB over existing methods.
format Article
id doaj-art-60bd7a57998a49a1a472720947e73518
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-60bd7a57998a49a1a472720947e735182025-08-20T03:28:52ZengIEEEIEEE Access2169-35362025-01-011311310.1109/ACCESS.2025.358315611050405MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View StereoByeonggwon Lee0https://orcid.org/0009-0006-7065-4263Junkyu Park1Khang Truong Giang2https://orcid.org/0000-0002-6709-7885Sungho Jo3https://orcid.org/0000-0002-7618-362XSoohwan Song4https://orcid.org/0000-0003-2145-1161Department of Computer Science and Artificial Intelligence, Dongguk University, Seoul, Republic of KoreaDepartment of Computer Science and Artificial Intelligence, Dongguk University, Seoul, Republic of Korea42dot, Seongnam-si, Gyeonggi-do, Republic of KoreaSchool of Computing, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Republic of KoreaDepartment of Computer Science and Artificial Intelligence, Dongguk University, Seoul, Republic of KoreaThis study addresses the challenge of online 3D model generation for neural rendering using an RGB image stream. Previous research has tackled this issue by incorporating Neural Radiance Fields (NeRF) or 3D Gaussian Splatting (3DGS) as scene representations within dense SLAM methods. However, most studies focus primarily on estimating coarse 3D scenes rather than achieving detailed reconstructions. Moreover, depth estimation based solely on images is often ambiguous, resulting in low-quality 3D models that lead to inaccurate renderings. To overcome these limitations, we propose a novel framework for high-quality 3DGS modeling that leverages an online multi-view stereo (MVS) approach. Our method estimates MVS depth using sequential frames from a local time window and applies comprehensive depth refinement techniques to filter out outliers. The refinement method produces temporally consistent depths by checking sequential geometric consistency, enabling accurate initialization of Gaussians in 3DGS. Furthermore, we introduce a parallelized backend module that optimizes the 3DGS model efficiently, ensuring timely updates with each new keyframe. Experimental results demonstrate that our method outperforms state-of-the-art dense SLAM methods, achieving an average PSNR improvement of approximately 2 dB on indoor scenes. Moreover, our method reliably produces consistent 3D models in complex outdoor scenes, where existing methods often fail due to tracking errors and depth noise. It also reconstructs large-scale aerial scenes effectively, achieving an average PSNR gain of about 10.28 dB over existing methods.https://ieeexplore.ieee.org/document/11050405/Online multi-view stereo3D Gaussian splattingneural renderingdense SLAM3D modelingdepth estimation
spellingShingle Byeonggwon Lee
Junkyu Park
Khang Truong Giang
Sungho Jo
Soohwan Song
MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
IEEE Access
Online multi-view stereo
3D Gaussian splatting
neural rendering
dense SLAM
3D modeling
depth estimation
title MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
title_full MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
title_fullStr MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
title_full_unstemmed MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
title_short MVS-GS: High-Quality 3D Gaussian Splatting Mapping via Online Multi-View Stereo
title_sort mvs gs high quality 3d gaussian splatting mapping via online multi view stereo
topic Online multi-view stereo
3D Gaussian splatting
neural rendering
dense SLAM
3D modeling
depth estimation
url https://ieeexplore.ieee.org/document/11050405/
work_keys_str_mv AT byeonggwonlee mvsgshighquality3dgaussiansplattingmappingviaonlinemultiviewstereo
AT junkyupark mvsgshighquality3dgaussiansplattingmappingviaonlinemultiviewstereo
AT khangtruonggiang mvsgshighquality3dgaussiansplattingmappingviaonlinemultiviewstereo
AT sunghojo mvsgshighquality3dgaussiansplattingmappingviaonlinemultiviewstereo
AT soohwansong mvsgshighquality3dgaussiansplattingmappingviaonlinemultiviewstereo