Robust stairway-detection and localization method for mobile robots using a graph-based model and competing initializations

Abstract

One of the major challenges for mobile robots in human-shaped environments is navigating stairways. This study presents a method for accurately detecting, localizing, and estimating the characteristics of stairways using point cloud data. The main challenge is the wide variety of different structures and shapes of stairways. This challenge is often aggravated by an unfavorable position of the sensor, which leaves large parts of the stairway occluded. This can be further aggravated by sparse point data. We overcome these difficulties by introducing a three-dimensional graph-based stairway-detection method combined with competing initializations. The stairway graph characterizes the general structural design of stairways in a generic way that can be used to describe a large variety of different stairways. By using multiple ways to initialize the graph, we can robustly detect stairways even if parts of the stairway are occluded. Furthermore, by letting the initializations compete against each other, we find the best initialization that accurately describes the measured stairway. The detection algorithm utilizes a plane-based approach. We also investigate different planar segmentation algorithms and experimentally compare them in an application-orientated manner. Our system accurately detects and estimates the stairway parameters with an average error of only 2:5 mm for a variety of stairways including ascending, descending, and spiral stairways. Our method works robustly with different depth sensors for either small-or large-scale environments and for dense and sparse point cloud data. Despite this generality, our system's accuracy is higher than most state-of-the-art stairway-detection methods.

Terrain-aware semantic mapping for cooperative subterranean exploration

“…While a geometric assessment of the environment prevents the robot from traversing over unsafe areas, there are situations where we want to exploit semantic information to explicitly traverse over known obstacles such as a stairway. To this end, we utilize the aforementioned unlabeled 3D pointcloud to classify stairways in the environment based on the work in Westfechtel et al, 2016 , Westfechtel et al, 2018 . We utilize this package in the proposed work for its superior performance in detecting various types of stairways from sparse pointcloud data.…”

“… Stairways (A) are mapped during a physical test deployment in the Engineering Center at the University of Colorado Boulder in Boulder, CO, United States of America. Parts of a 3D pointcloud that belong to a stairway white in (B) are segmented via StairwayDetector ( Westfechtel et al, 2016 ; 2018 ) and then fused into an occupancy map (C) . Additionally, we cluster stair voxels (blue) and extract the primary axis pink in (D) to inform the terrain-aware planner.…”

“… Sánchez-Rojas et al (2021) presents a hybrid visual and geometric approach to stair detection as well as a fuzzy logic controller for the alignment of robots to the detected stairway. ( Westfechtel et al, 2016 , Westfechtel et al, 2018 ) introduces StairwayDetection, an open-source package that directly classifies the various parts of stairways using a graph-based approach, including the stair tread, riser, and railing segments.…”

Flexible Supervised Autonomy for Exploration in Subterranean Environments

“…Semantic information on stairs is fused into the mapping framework using the open source StairwayDetection (Westfechtel et al, 2018) package and a binary Bayes filter (Thrun et al, 2005). Stair classification of point clouds via this approach consists of 4 major steps: (1) pre-analysis, in which the point cloud is downsampled and filtered, normal and curvature is estimated for each point, and floor separation is performed;…”

Staircase Detection, Characterization and Approach Pipeline for Search and Rescue Robots

“…Once this is done, the staircase bottom mid-point is calculated. Finally in [18], they present a complete pipeline to detect, localize and estimate the characteristics of the staircase using point cloud data obtained from a LIDAR. For detection, they use a plane-based approach and later estimate the staircase parameters with an error of only 2.5 mm.…”