3D Dense Mapping with the Graph of Keyframe-Based and View-Dependent Local Maps

This article concerns the problem of a dense mapping system for a robot exploring a new environment. In this scenario, a robot equipped with an RGB-D camera uses RGB and range data to build a consistent model of the environment. Firstly, dense mapping requires the selection of the data representation. Secondly, the dense mapping system has to deal with localization drift which can be corrected when loop closure is detected. In this article, we deal with both of these problems, and we make several technical contributions. We define local maps which use the Normal Distribution Transform (NDT) stored in the 2D structures to represent the local scene with varying 3D resolution. This method directly utilizes the uncertainty model of the range sensor and provides information about the accuracy of the data in the map. We also propose an architecture that utilizes pose and covisibility graphs to correct a global model of the environment after loop closure detection. We show how to integrate the dense local mapping with the pose graph and keyframes management system in the ORB-SLAM2 localization. Finally, we show the advantages of the view-dependent model over the methods that uniformly divide the space to represent objects in the environment.

Traversability Assessment and Trajectory Planning of Unmanned Ground Vehicles with Suspension Systems on Rough Terrain

This paper presents a traversability assessment method and a trajectory planning method. They are key features for the navigation of an unmanned ground vehicle (UGV) in a non-planar environment. In this work, a 3D light detection and ranging (LiDAR) sensor is used to obtain the geometric information about a rough terrain surface. For a given SE(2) pose of the vehicle and a specific vehicle model, the SE(3) pose of the vehicle is estimated based on LiDAR points, and then a traversability is computed. The traversability tells the vehicle the effects of its interaction with the rough terrain. Note that the traversability is computed on demand during trajectory planning, so there is not any explicit terrain discretization. The proposed trajectory planner finds an initial path through the non-holonomic A*, which is a modified form of the conventional A* planner. A path is a sequence of poses without timestamps. Then, the initial path is optimized in terms of the traversability, using the method of Lagrange multipliers. The optimization accounts for the model of the vehicle's suspension system. Therefore, the optimized trajectory is dynamically feasible, and the trajectory tracking error is small. The proposed methods were tested in both the simulation and the real-world experiments. The simulation experiments were conducted in a simulator called Gazebo, which uses a physics engine to compute the vehicle motion. The experiments were conducted in various non-planar experiments. The results indicate that the proposed methods could accurately estimate the SE(3) pose of the vehicle. Besides, the trajectory cost of the proposed planner was lower than the trajectory costs of other state-of-the-art trajectory planners.

Heterogeneous Map Merging: State of the Art

Multi-robot mapping and environment modeling have several advantages that makeit an attractive alternative to the mapping with a single robot: faster exploration, higherfault tolerance, richer data due to different sensors being used by different systems. However,the environment modeling with several robotic systems operating in the same area causes problemsof higher-order—acquired knowledge fusion and synchronization over time, revealing the sameenvironment properties using different sensors with different technical specifications. While theexisting robot map and environment model merging techniques allow merging certain homogeneousmaps, the possibility to use sensors of different physical nature and different mapping algorithms islimited. The resulting maps from robots with different specifications are heterogeneous, and eventhough some research on how to merge fundamentally different maps exists, it is limited to specificapplications. This research reviews the state of the art in homogeneous and heterogeneous mapmerging and illustrates the main research challenges in the area. Six factors are identified thatinfluence the outcome of map merging: (1) robotic platform hardware configurations, (2) maprepresentation types, (3) mapping algorithms, (4) shared information between robots, (5) relativepositioning information, (6) resulting global maps.

Learning the Cost Function for Foothold Selection in a Quadruped Robot

This paper is focused on designing a cost function of selecting a foothold for a physical quadruped robot walking on rough terrain. The quadruped robot is modeled with Denavit–Hartenberg (DH) parameters, and then a default foothold is defined based on the model. Time of Flight (TOF) camera is used to perceive terrain information and construct a 2.5D elevation map, on which the terrain features are detected. The cost function is defined as the weighted sum of several elements including terrain features and some features on the relative pose between the default foothold and other candidates. It is nearly impossible to hand-code the weight vector of the function, so the weights are learned using Supporting Vector Machine (SVM) techniques, and the training data set is generated from the 2.5D elevation map of a real terrain under the guidance of experts. Four candidate footholds around the default foothold are randomly sampled, and the expert gives the order of such four candidates by rotating and scaling the view for seeing clearly. Lastly, the learned cost function is used to select a suitable foothold and drive the quadruped robot to walk autonomously across the rough terrain with wooden steps. Comparing to the approach with the original standard static gait, the proposed cost function shows better performance.

A 2.5D Map-Based Mobile Robot Localization via Cooperation of Aerial and Ground Robots

Recently, there has been increasing interest in studying the task coordination of aerial and ground robots. When a robot begins navigation in an unknown area, it has no information about the surrounding environment. Accordingly, for robots to perform tasks based on location information, they need a simultaneous localization and mapping (SLAM) process that uses sensor information to draw a map of the environment, while simultaneously estimating the current location of the robot on the map. This paper aims to present a localization method based in cooperation between aerial and ground robots in an indoor environment. The proposed method allows a ground robot to reach accurate destination by using a 2.5D elevation map built by a low-cost RGB-D (Red Green and Blue-Depth) sensor and 2D Laser sensor attached onto an aerial robot. A 2.5D elevation map is formed by projecting height information of an obstacle using depth information obtained by the RGB-D sensor onto a grid map, which is generated by using the 2D Laser sensor and scan matching. Experimental results demonstrate the effectiveness of the proposed method for its accuracy in location recognition and computing speed.

Occupancy-elevation grid: an alternative approach for robotic mapping and navigation

This paper proposes an alternative environment mapping method for accurate robotic navigation based on 3D information. Typical techniques for 3D mapping using occupancy grid require intensive computational workloads in order to both build and store the map. This work introduces an Occupancy-Elevation Grid (OEG) mapping technique, which is a discrete mapping approach where each cell represents the occupancy probability, the height of the terrain and its variance. This representation allows a mobile robot to know with an accurate degree of certainty whether a place in the environment is occupied by an obstacle and the height of such obstacle. Thus, based on its hardware characteristics, it can make calculations to decide if it is possible to traverse that specific place. In general, the map representation introduced can be used in conjunction with any kind of distance sensor. In this work, we use laser range data and stereo system data with a probabilistic treatment. The resulting maps allow the execution of tasks as decision making for autonomous navigation, exploration, localization and path planning, considering the existence and the height of the obstacles. Experiments carried out with real data demonstrate that the proposed approach yields useful maps for autonomous navigation.

Detection of traversable region around a mobile robot by computing terrain unevenness from the range data of a 3D laser scanner

Purpose

– The purpose of this paper is to detect traversable regions surrounding a mobile robot by computing terrain unevenness using the range data obtained from a single 3D scan.

Design/methodology/approach

– The geometry of acquiring range data from a 3D scan is exploited to probe the terrain and extract traversable regions. Nature of terrain under each scan point is quantified in terms of an unevenness value, which is computed from the difference in range of scan point with respect to its neighbours. Both radial and transverse unevenness values are computed and compared with threshold values at every point to determine if the point belongs to a traversable region or an obstacle. A region growing algorithm spreads like a wavefront to join all traversable points into a traversable region.

Findings

– This simple method clearly distinguishes ground and obstacle points. The method works well even in presence of terrain slopes or when the robot experiences pitch and roll.

Research limitations/implications

– The method applies on single 3D scans and not on aggregated point cloud in general.

Practical implications

– The method has been tested on a mobile robot in outdoor environment in our research centre.

Social implications

– This method, along with advanced navigation schemes, can reduce human intervention in many mobile robot applications including unmanned ground vehicles.

Originality/value

– Range difference between scan points has been used earlier for obstacle detection, but no methodology has been developed around this concept. The authors propose a concrete method based on computation of radial and transverse unevenness at every point and detecting obstacle edges using range-dependent threshold values.

A Set-Theoretic Algorithm for Real-Time Terrain Mapping of Mobile Robots in Outdoor Environments

In this paper, we address the problem of mapping outdoor rough terrain environments for mobile robots. While uncertainties arising from multiple sources are considered explicitly and assumed to be unknown-but-bounded, a set-theoretic framework is proposed to construct the terrain model as a set-valued elevation map that extends the notion of the elevation map with elevation variation in each cell stored by intervals. The localization problem of the mobile robot is also considered and solved by a set-membership filter in order to provide guaranteed bounded-pose estimation, which can be incorporated to the elevation map to improve the accuracy of the final terrain model. A more compact terrain representation can be obtained by the proposed algorithm with relatively low computational complexity, which makes it suitable for real-time applications. Furthermore, improved smoothness is achieved by the inherent conservativeness of the set-theoretic method without additional filtering or interpolation processes. Simulations as well as real-life experiments of a mobile robot operating in outdoor rough terrain environments with a 2D scanning laser rangefinder demonstrate the effectiveness and robustness of the proposed method.

Terrain Inclination Aided Three-Dimensional Localization and Mapping for an Outdoor Mobile Robot

A new 3D localization and mapping technique with terrain inclination assistance is proposed in this paper to allow a robot to identify its location and build a global map in an outdoor environment. The Iterative Closest Points (ICP) algorithm and terrain inclination-based localization are combined together to achieve accurate and fast localization and mapping. Inclinations of the terrains the robot navigates are used to achieve local localization during the interval between two laser scans. Using the results of the above localization as the initial condition, the ICP algorithm is then applied to align the overlapped laser scan maps to update the overhanging obstacles for building a global map of the surrounding area. Comprehensive experiments were carried out for the validation of the proposed 3D localization and mapping technique. The experimental results show that the proposed technique could reduce time consumption and improve the accuracy of the performance.

Probabilistic multi-level maps from LIDAR data

Recent research has shown that robots can model their world with Multi-Level (ML) maps, which utilize patches in a two-dimensional grid space to represent various environment elevations within a given grid cell. Although these maps are able to produce three-dimensional models of the environment while exploiting the computational feasibility of single elevation maps, they do not take into account in-plane uncertainty when matching measurements to grid cells or when grouping those measurements into patches. To respond to these drawbacks, this paper proposes to extend these ML maps into Probabilistic Multi-Level (PML) maps, which use formal probability theory to incorporate estimation and modeling errors due to uncertainty. Measurements are probabilistically associated with cells near the nominal location, and are categorized through hypothesis testing into patches via classification methods that incorporate uncertainty. Experimental results on representative objects found in both indoor and outdoor environments show that PML generally outperforms ML, including in noisy and sparse data environments, by producing more consistent, informative and conservative maps. In addition, PML provides the framework to heterogeneous, cooperative mapping and a way to probabilistically discriminate between conflicting maps.

Space-carving Kernels for Accurate Rough Terrain Estimation

Accurate terrain estimation is critical for autonomous offroad navigation. Reconstruction of a three-dimensional (3D) surface allows rough and hilly ground to be represented, yielding faster driving and better planning and control. However, data from a 3D sensor samples the terrain unevenly, quickly becoming sparse at longer ranges and containing large voids because of occlusions and inclines. The proposed approach uses online kernel-based learning to estimate a continuous surface over the area of interest while providing upper and lower bounds on that surface. Unlike other approaches, visibility information is exploited to constrain the terrain surface and increase precision, and an efficient gradient-based optimization allows for realtime implementation. To model sensor noise over varying ranges, a non-stationary covariance function is adopted. Experimental results are presented for several datasets, including groundtruthed terrain and a large 3D stereo dataset.