FloorNet: A Unified Framework for Floorplan Reconstruction from 3D Scans

Abstract

The ultimate goal of this indoor mapping research is to automatically reconstruct a floorplan simply by walking through a house with a smartphone in a pocket. This paper tackles this problem by proposing FloorNet, a novel deep neural architecture. The challenge lies in the processing of RGBD streams spanning a large 3D space. FloorNet effectively processes the data through three neural network branches: 1) PointNet with 3D points, exploiting the 3D information; 2) CNN with a 2D point density image in a top-down view, enhancing the local spatial reasoning; and 3) CNN with RGB images, utilizing the full image information. FloorNet exchanges intermediate features across the branches to exploit the best of all the architectures. We have created a benchmark for floorplan reconstruction by acquiring RGBD video streams for 155 residential houses or apartments with Google Tango phones and annotating complete floorplan information. Our qualitative and quantitative evaluations demonstrate that the fusion of three branches effectively improves the reconstruction quality. We hope that the paper together with the benchmark will be an important step towards solving a challenging vector-graphics reconstruction problem. Code and data are available at https://github.com/art-programmer/FloorNet.

Floor-SP: Inverse CAD for Floorplans by Sequential Room-Wise Shortest Path

“…Fig. 6 compares Floor-SP against the current state-ofthe-art FloorNet [20] and the variants of our system. Floor-Net follows a bottom-up process, where it first detects corners then uses Integer Programming to find their valid connections.…”

“…While many previous works utilize RGBD scans/point clouds for high-quality indoor reconstruction [17,19,23,20], FloorNet [20] is the current state-of-the-art for floorplan reconstruction task tested on large-scale indoor benchmarks. FloorNet combines DNN and IP in a bottom-up process but it has three major failure modes.…”

“…Lastly, a large-scale synthetic dataset, SUNCG [26], offers a variety of indoor scenes. For the floorplan reconstruction task, FloorNet [20] provides the benchmark with full floorplan annotations and the corresponding RGBD videos from smartphones for 155 residential units. This paper utilizes production-quality panorama RGBD scans for 527 houses or apartments with floorplan annotations.…”

“…The aligned panorama RGBD scans are first converted into 2D point-density/normal map, which is the input to Floor-SP. Unlike FloorNet [20], we focus on the wall structures, where doors/windows, icons, and room semantics can be added given proper wall structures.…”

“…We have evaluated the proposed approach on productionquality RGBD scans of 527 apartments or houses, a few times larger than the current largest database [20]. Our approach makes significant improvements over the current state-of-the-art [20]. We refer to our project website http: //jcchen.me/floor-sp/ for code and data.…”