Automatic recognition and segmentation of architectural elements from 2D drawings by convolutional neural network

Yahan Xiao, Sen Chen, Yasushi Ikeda, Kensuke Hotta

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The BIM modeling process is the most time-consuming aspect. This paper studies the possibility of applying the recognition and segmentation of architectural components by deep learning to assist automatic BIM modeling. The research has two parts: the first one is dataset preparing, that images with the labeled architectural components from an original CAD drawing are made for the network training, and second is training and testing, that a mature network which has been trained in hundreds of labeled images is used to make predictions. The utilization of the current study results is discussed and the optimization method as well.

Original languageEnglish
Title of host publicationRE
Subtitle of host publicationAnthropocene, Design in the Age of Humans - Proceedings of the 25th International Conference on Computer-Aided Architectural Design Research in Asia, CAADRIA 2020
EditorsDominik Holzer, Walaiporn Nakapan, Anastasia Globa, Immanuel Koh
PublisherThe Association for Computer-Aided Architectural Design Research in Asia (CAADRIA)
Pages843-852
Number of pages10
ISBN (Electronic)9789887891734
Publication statusPublished - 2020
Event25th International Conference on Computer-Aided Architectural Design Research in Asia, CAADRIA 2020 - Bangkok, Thailand
Duration: 2020 Aug 52020 Aug 6

Publication series

NameRE: Anthropocene, Design in the Age of Humans - Proceedings of the 25th International Conference on Computer-Aided Architectural Design Research in Asia, CAADRIA 2020
Volume1

Conference

Conference25th International Conference on Computer-Aided Architectural Design Research in Asia, CAADRIA 2020
CountryThailand
CityBangkok
Period20/8/520/8/6

Keywords

  • BIM
  • CAD drawings
  • Computer vision
  • Convolutional Neural Network
  • Recognition and Segmentation

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Building and Construction

Fingerprint Dive into the research topics of 'Automatic recognition and segmentation of architectural elements from 2D drawings by convolutional neural network'. Together they form a unique fingerprint.

Cite this