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

Abstract

Figure 1. The proposed system, dubbed Floor-SP, takes aligned panorama RGBD scans as input, finds room segments, solves an optimization problem to reconstruct a floorplan graph as multiple polygonal loops (one for each room), and merges them into a 2D graph via simple post-processing heuristics. The optimization is the technical contribution of the paper, which employs the room-wise coordinate descent strategy and sequentially solves shortest path problems to optimize the room structure. AbstractThis paper proposes a new approach for automated floorplan reconstruction from RGBD scans, a major milestone in indoor mapping research. The approach, dubbed Floor-SP, formulates a novel optimization problem, where room-wise coordinate descent sequentially solves shortest path problems to optimize the floorplan graph structure. The objective function consists of data terms guided by deep neural networks, consistency terms encouraging adjacent rooms to share corners and walls, and the model complexity term. The approach does not require corner/edge primitive extraction unlike most other methods. We have evaluated our system on production-quality RGBD scans of 527 apartments or houses, including many units with non-Manhattan structures. Qualitative and quantitative evaluations demonstrate a significant performance boost over the current state-of-the-art. Please refer to our project website

How Many Bedrooms Do You Need? A Real-Estate Recommender System from Architectural Floor Plan Images

“…Recently, an improved method that can generate the floor plan via 3D scanning with higher accuracy and higher speed is introduced by [10]. Specifically, the proposed pipeline, namely, Scan2Plan demonstrated outstanding results when evaluated on the publicly available Structured3D [11] and BKE [12] datasets.…”

A Comparison of Low-Cost Sensor Systems in Automatic Cloud-Based Indoor 3D Modeling

“…Objects further than 10 cm from the lens are visible independent of their position with respect to the lenses [50]. The camera can be controlled either by an on-camera user interface or through a mobile application [51], with the latter being used for Matterport processing [38].…”

“…Alternatively, a collection of range and color maps in a reference frame can be used as input for the reconstruction, with the sensor's positioning information aiding reconstruction [30]. For modeling RGB-D data, deep learning approaches are common, with methods including the integration of the 2D semantic segmentation of images with a 3D reconstruction [36], the 2D reconstruction of walls as polygonal loops [37], and 3D volumetric segmentation [38].…”

Floorplan generation from 3D point clouds: A space partitioning approach

“…Reconstructing the floorplan from point clouds is usually operated in two steps. First, some geometric primitives are detected from input 3D data either by traditional primitive detection methods [3,4] or through learning approaches [5,6]. These primitives locally describe the position of walls by planes or room corners by points, which form basic elements to represent the geometry of each room.…”