Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds

Representation of wall facade in 3D scene is important for scene understanding and reconstruction. Doors and windows are essential elements of wall facade, which enrich the models of indoor buildings. In order to reconstruct the wall facade with detailed wall opening (doors and windows) informatio...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	NING, X., WANG, M., TANG, J., ZHANG, H., WANG, Y.
Formato:	article
Lenguaje:	EN
Publicado:	Stefan cel Mare University of Suceava 2021
Materias:	computer graphics computer aided analysis feature extraction object segmentation pattern recognition Electrical engineering. Electronics. Nuclear engineering TK1-9971 Computer engineering. Computer hardware TK7885-7895
Acceso en línea:	https://doaj.org/article/88a588d1fd62483aaa547114be8e743c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:88a588d1fd62483aaa547114be8e743c
record_format	dspace
spelling	oai:doaj.org-article:88a588d1fd62483aaa547114be8e743c2021-12-05T17:03:49ZStructural Wall Facade Reconstruction of Scanned Scene in Point Clouds1582-74451844-760010.4316/AECE.2021.04002https://doaj.org/article/88a588d1fd62483aaa547114be8e743c2021-11-01T00:00:00Zhttp://dx.doi.org/10.4316/AECE.2021.04002https://doaj.org/toc/1582-7445https://doaj.org/toc/1844-7600Representation of wall facade in 3D scene is important for scene understanding and reconstruction. Doors and windows are essential elements of wall facade, which enrich the models of indoor buildings. In order to reconstruct the wall facade with detailed wall opening (doors and windows) information from point cloud, an improved automatic reconstruction approach is proposed in this paper. Specifically, we first detect the occlusion area on the wall by ray-tracing method. Then, the interested wall facade is segmented from scenes, and the contour and feature lines of the wall surface are extracted through the analysis of spatial occupancy. Third, the wall point cloud is transformed into a 2D cell complex. The wall surface is finally reconstructed through the wall opening identification based on graph cut. Experimental results demonstrate that our method has better performance in detail description, feature line extraction and opening area detection of the scene point cloud than existing state-of-the-art methods.NING, X.WANG, M.TANG, J.ZHANG, H.WANG, Y.Stefan cel Mare University of Suceavaarticlecomputer graphicscomputer aided analysisfeature extractionobject segmentationpattern recognitionElectrical engineering. Electronics. Nuclear engineeringTK1-9971Computer engineering. Computer hardwareTK7885-7895ENAdvances in Electrical and Computer Engineering, Vol 21, Iss 4, Pp 11-20 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	computer graphics computer aided analysis feature extraction object segmentation pattern recognition Electrical engineering. Electronics. Nuclear engineering TK1-9971 Computer engineering. Computer hardware TK7885-7895
spellingShingle	computer graphics computer aided analysis feature extraction object segmentation pattern recognition Electrical engineering. Electronics. Nuclear engineering TK1-9971 Computer engineering. Computer hardware TK7885-7895 NING, X. WANG, M. TANG, J. ZHANG, H. WANG, Y. Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds
description	Representation of wall facade in 3D scene is important for scene understanding and reconstruction. Doors and windows are essential elements of wall facade, which enrich the models of indoor buildings. In order to reconstruct the wall facade with detailed wall opening (doors and windows) information from point cloud, an improved automatic reconstruction approach is proposed in this paper. Specifically, we first detect the occlusion area on the wall by ray-tracing method. Then, the interested wall facade is segmented from scenes, and the contour and feature lines of the wall surface are extracted through the analysis of spatial occupancy. Third, the wall point cloud is transformed into a 2D cell complex. The wall surface is finally reconstructed through the wall opening identification based on graph cut. Experimental results demonstrate that our method has better performance in detail description, feature line extraction and opening area detection of the scene point cloud than existing state-of-the-art methods.
format	article
author	NING, X. WANG, M. TANG, J. ZHANG, H. WANG, Y.
author_facet	NING, X. WANG, M. TANG, J. ZHANG, H. WANG, Y.
author_sort	NING, X.
title	Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds
title_short	Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds
title_full	Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds
title_fullStr	Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds
title_full_unstemmed	Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds
title_sort	structural wall facade reconstruction of scanned scene in point clouds
publisher	Stefan cel Mare University of Suceava
publishDate	2021
url	https://doaj.org/article/88a588d1fd62483aaa547114be8e743c
work_keys_str_mv	AT ningx structuralwallfacadereconstructionofscannedsceneinpointclouds AT wangm structuralwallfacadereconstructionofscannedsceneinpointclouds AT tangj structuralwallfacadereconstructionofscannedsceneinpointclouds AT zhangh structuralwallfacadereconstructionofscannedsceneinpointclouds AT wangy structuralwallfacadereconstructionofscannedsceneinpointclouds
_version_	1718371309517275136

Structural Wall Facade Reconstruction of Scanned Scene in Point Clouds

Ejemplares similares