Piecewise‐planar Reconstruction of Multi‐room Interiors with Arbitrary Wall Arrangements

Abstract

Figure 1: Given a set of cluttered 3D point clouds of building interiors (LEFT), we use fitting rectangles of the planar components as simplified scene description, and separate those belonging to permanent structures (green) from clutter (red) (MIDDLE). From the dominant planes of the permanent components we build a 3D cell complex, whose cells are partitioned to create individual room polyhedra (RIGHT). AbstractReconstructing the as-built architectural shape of building interiors has emerged in recent years as an important and challenging research problem. An effective approach must be able to faithfully capture the architectural structures and separate permanent components from clutter (e.g. furniture), while at the same time dealing with defects in the input data. For many applications, higher-level information on the environment is also required, in particular the shape of individual rooms. To solve this ill-posed problem, state-of-the-art methods assume constrained input environments with a 2.5D or, more restrictively, a Manhattan-world structure, which significantly restricts their applicability in real-world settings. We present a novel pipeline that allows to reconstruct general 3D interior architectures, significantly increasing the range of real-world architectures that can be reconstructed and labeled by any interior reconstruction method to date. Our method finds candidate permanent components by reasoning on a graph-based scene representation, then uses them to build a 3D linear cell complex that is partitioned into separate rooms through a multi-label energy minimization formulation. We demonstrate the effectiveness of our method by applying it to a variety of real-world and synthetic datasets and by comparing it to more specialized state-of-the-art approaches.

Exploiting Indoor Mobile Laser Scanner Trajectories for Semantic Interpretation of Point Clouds

“…In their model, topological relations between spaces and rooms are reconstructed in a graph structure, which simplifies to update the model in the future. In a recent work by Mura et al (2016), the authors encode the indoor components in an adjacency graph and -by finding the path from ceiling to floor -try to detect and reconstruct the main components. Armeni et al (2016) parse point clouds collected with a Matterport system into indoor components for large-scale indoor scenes.…”

“…Partitioning the space to meaningful areas such as rooms and corridors is referred to room segmentation, space subdivision or space partitioning (Krūminaitė and Zlatanova, 2014;Mura et al, 2016;Turner and Zakhor, 2014). By using mobile laser scanners, there is no information for the partitioning to individual rooms because we are not aware which points are captured in the same room.…”

Reconstruction of Three-Dimensional (3D) Indoor Interiors with Multiple Stories via Comprehensive Segmentation

“…Moreover, rebuilding an indoor watertight model is difficult in certain cases [26,27]. In recent approaches, the focus shifted towards segmentation into different individual rooms to resolve issues with missing data for ceilings and floors [14], and then reconstruct indoor models through segmented maps.…”

“…Laser-scanning technologies are significantly advanced [4,[10][11][12] and they can rapidly capture the details of a complex indoor structure's geometry; thus, laser-scanning technologies show promise for certain applications [8]. The high-quality reconstruction of a watertight mesh model from scanned studied [1,14], although the addressed issues are different. Early research on indoor reconstruction focused on detecting planes to reconstruct indoor environments, similar to methods of outdoor reconstruction.…”

The Isprs Benchmark on Indoor Modelling

“…In practice, most existing methods reconstruct geometric elements only (Becker et al, 2015;Díaz-Vilariño et al, 2015;Mura et al, 2016;Oesau et al, 2014;Sanchez and Zakhor, 2012;Thomson and Boehm, 2015;Valero et al, 2012;Xiao and Furukawa, 2012;Xiong et al, 2013), while a few can model spaces and the topological relations between them (Khoshelham and Díaz-Vilariño, 2014;Tran et al, 2017). Nonetheless, a comprehensive evaluation should take into account all three components of an indoor model.…”