Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding

Indoor scenes often contain complex layouts and interactions between objects, making 3D modeling of point clouds inherently difficult. In this paper, we design a divide-and-conquer modeling method considering the structural differences between indoor walls and internal objects. To achieve semantic u...

Full description

Saved in:
Bibliographic Details
Main Authors: Minglei Li, Mingfan Li, Min Li, Leheng Xu
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/4/596
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850081986423554048
author Minglei Li
Mingfan Li
Min Li
Leheng Xu
author_facet Minglei Li
Mingfan Li
Min Li
Leheng Xu
author_sort Minglei Li
collection DOAJ
description Indoor scenes often contain complex layouts and interactions between objects, making 3D modeling of point clouds inherently difficult. In this paper, we design a divide-and-conquer modeling method considering the structural differences between indoor walls and internal objects. To achieve semantic understanding, we propose an effective 3D instance segmentation module using a deep network Indoor3DNet combined with super-point clustering, which provides a larger receptive field and maintains the continuity of individual objects. The Indoor3DNet includes an efficient point feature extraction backbone with good operability for different object granularity. In addition, we use a geometric primitives-based modeling approach to generate lightweight polygonal facets for walls and use a cross-modal registration technique to fit the corresponding instance models for internal objects based on their semantic labels. This modeling method can restore correct geometric shapes and topological relationships while maintaining a very lightweight structure. We have tested the method on diverse datasets, and the experimental results demonstrate that the method outperforms the state-of-the-art in terms of performance and robustness.
format Article
id doaj-art-fe377cbdc5d54ca485791cf34ec86556
institution DOAJ
issn 2072-4292
language English
publishDate 2025-02-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-fe377cbdc5d54ca485791cf34ec865562025-08-20T02:44:36ZengMDPI AGRemote Sensing2072-42922025-02-0117459610.3390/rs17040596Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene UnderstandingMinglei Li0Mingfan Li1Min Li2Leheng Xu3College of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaIndoor scenes often contain complex layouts and interactions between objects, making 3D modeling of point clouds inherently difficult. In this paper, we design a divide-and-conquer modeling method considering the structural differences between indoor walls and internal objects. To achieve semantic understanding, we propose an effective 3D instance segmentation module using a deep network Indoor3DNet combined with super-point clustering, which provides a larger receptive field and maintains the continuity of individual objects. The Indoor3DNet includes an efficient point feature extraction backbone with good operability for different object granularity. In addition, we use a geometric primitives-based modeling approach to generate lightweight polygonal facets for walls and use a cross-modal registration technique to fit the corresponding instance models for internal objects based on their semantic labels. This modeling method can restore correct geometric shapes and topological relationships while maintaining a very lightweight structure. We have tested the method on diverse datasets, and the experimental results demonstrate that the method outperforms the state-of-the-art in terms of performance and robustness.https://www.mdpi.com/2072-4292/17/4/5963D indoor modelssemantic segmentationpoint cloud clusteringIndoor3DNet
spellingShingle Minglei Li
Mingfan Li
Min Li
Leheng Xu
Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding
Remote Sensing
3D indoor models
semantic segmentation
point cloud clustering
Indoor3DNet
title Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding
title_full Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding
title_fullStr Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding
title_full_unstemmed Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding
title_short Building Lightweight 3D Indoor Models from Point Clouds with Enhanced Scene Understanding
title_sort building lightweight 3d indoor models from point clouds with enhanced scene understanding
topic 3D indoor models
semantic segmentation
point cloud clustering
Indoor3DNet
url https://www.mdpi.com/2072-4292/17/4/596
work_keys_str_mv AT mingleili buildinglightweight3dindoormodelsfrompointcloudswithenhancedsceneunderstanding
AT mingfanli buildinglightweight3dindoormodelsfrompointcloudswithenhancedsceneunderstanding
AT minli buildinglightweight3dindoormodelsfrompointcloudswithenhancedsceneunderstanding
AT lehengxu buildinglightweight3dindoormodelsfrompointcloudswithenhancedsceneunderstanding