Traversability-aware Consistent Situational Graphs for Indoor Localization and Mapping

TL;DR

The paper tackles inconsistent room recognition in indoor SLAM with scene graphs caused by restricted viewpoints and time-varying segmentation. It introduces a traversability-aware, hierarchical situational graph framework with a front-end that extracts traversable ground points, free-space clusters, and rooms, and a back-end that jointly optimizes the graph. Core contributions include a traversable ground-point extraction approach, $\lambda_{th}$-based free-space clustering, and a room-extraction strategy that flushes lower-layer data and leverages aisle-width changes to define meaningful rooms; these yield more coherent room objects and reduce PGO load. Experiments on the TIERS dataset show improved room re-detection, higher Dice coefficients, and lower ATE and $T_{PGO}$, demonstrating practical gains in reliable indoor localization and mapping.

Abstract

Scene graphs enhance 3D mapping capabilities in robotics by understanding the relationships between different spatial elements, such as rooms and objects. Recent research extends scene graphs to hierarchical layers, adding and leveraging constraints across these levels. This approach is tightly integrated with pose-graph optimization, improving both localization and mapping accuracy simultaneously. However, when segmenting spatial characteristics, consistently recognizing rooms becomes challenging due to variations in viewpoints and limited field of view (FOV) of sensors. For example, existing real-time approaches often over-segment large rooms into smaller, non-functional spaces that are not useful for localization and mapping due to the time-dependent method. Conversely, their voxel-based room segmentation method often under-segment in complex cases like not fully enclosed 3D space that are non-traversable for ground robots or humans, leading to false constraints in pose-graph optimization. We propose a traversability-aware room segmentation method that considers the interaction between robots and surroundings, with consistent feasibility of traversability information. This enhances both the semantic coherence and computational efficiency of pose-graph optimization. Improved performance is demonstrated through the re-detection frequency of the same rooms in a dataset involving repeated traversals of the same space along the same path, as well as the optimization time consumption.

Figures (3)

• Figure 1: Our proposed hierarchical graph layers. From the bottom, the hierarchical graph consists of traversable nodes, keyframes, walls, rooms, and floors layers.
• Figure 2: Overview of our method. Inputs are the 3D LiDAR measurements, which is pre-filtered while calculating traversability, and robot odometry. Front-end is to extract wall planes, rooms, floors and traversable nodes. and the graphs are jointly optimized in the back-end. We added traversability-aware room segmentation strategy for consistency.
• Figure 3: (L-R) Examples of complex corridor environments: (L) An open-corridor structure that shares 3D free space, where ground robots or humans cannot traverse. (R) Bird's-eye view of the global point cloud map of the 3rd floor data from the Indoor sequence of the TIERS dataset, which traverses through open-corridor environments sier2023benchmark.