Reality Capture and Digital Twins

Residential floor plan recognition and reconstruction

Xiaolei Lv, Shengchu Zhao, Xinyang Yu, Binqiang Zhao

Published 2021

Thumbnail for Residential floor plan recognition and reconstruction
📄 Paper
Recognition and reconstruction of residential floor plan drawings are important and challenging in design, decoration, and architectural remodeling fields. An automatic framework is provided that accurately recognizes the structure, type, and size of the room, and outputs vectorized 3D reconstruction results. Deep segmentation and detection neural networks are utilized to extract room structural information. Key points detection network and cluster analysis are utilized to calculate scales of rooms. The vectorization of room information is processed through an iterative optimization-based method. The system significantly increases accuracy and generalization ability, compared with existing methods. It outperforms other systems in floor plan segmentation and vectorization process, especially inclined wall detection.

Abstract

Recognition and reconstruction of residential floor plan drawings are important and challenging in design, decoration, and architectural remodeling fields. An automatic framework is provided that accurately recognizes the structure, type, and size of the room, and outputs vectorized 3D reconstruction results. Deep segmentation and detection neural networks are utilized to extract room structural information. Key points detection network and cluster analysis are utilized to calculate scales of rooms. The vectorization of room information is processed through an iterative optimization-based method. The system significantly increases accuracy and generalization ability, compared with existing methods. It outperforms other systems in floor plan segmentation and vectorization process, especially inclined wall detection.

Topics

2D floor plan3D floor planBIMFloor Plan Reconstruction and Parametrizationdigital twinsimage-to-planlayout recoveryscan-to-BIM

Cite this paper

@INPROCEEDINGS{d78d729a,
  author={Lv, Xiaolei and Zhao, Shengchu and Yu, Xinyang and Zhao, Binqiang},
  booktitle={2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  title={Residential floor plan recognition and reconstruction},
  year={2021},
  volume={},
  number={},
  pages={16712-16721},
  keywords={Computer vision;Three-dimensional displays;Neural networks;Benchmark testing;Iterative methods;Data mining;Floors},
  doi={10.1109/CVPR46437.2021.01644}}