Repository logo
 

DEEP-LEARNING GUIDED STRUCTURAL OBJECT DETECTION IN LARGE-SCALE, OCCLUDED INDOOR POINT CLOUD DATASETS

Accepted version
Peer-reviewed

Type

Conference Object

Change log

Authors

Drobnyi, V 
Li, S 

Abstract

Automatic geometry digitisation of existing buildings remains challenging due to the large scale and heavy clutter of input point clouds. This paper presents a twostage hybrid method to detect structural objects. The first stage detects areas of interest that are likely to contain an object, while the second stage finds precise objects. The method benefits from data-driven and model-driven approaches to achieve high accuracy for large-scale, highly cluttered and occluded real-world environments. We evaluate our method on the Stanford3D S3DIS dataset to show that the method detects from 83% to 98% of structural objects, such as columns, doors and windows.

Description

Keywords

4013 Geomatic Engineering, 46 Information and Computing Sciences, 40 Engineering

Journal Title

Proceedings of the European Conference on Computing in Construction

Conference Name

2023 European Conference on Computing in Construction and the 40th International CIB W78 Conference

Journal ISSN

2684-1150
2684-1150

Volume Title

4

Publisher

European Council for Computing in Construction
Sponsorship
European Commission Horizon 2020 (H2020) Marie Sk?odowska-Curie actions (860555)
European Commission Horizon 2020 (H2020) Industrial Leadership (IL) (958398)