Online LiDAR-SLAM for Legged Robots with Robust Registration and Deep-Learned Loop Closure

Ramezani, Milad; Tinchev, Georgi; Egor, Iuganov; Fallon, Maurice

doi:10.1109/icra40945.2020.9196769

Cited by 33 publications

(16 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Kinematic-Inertial Odometry is effective for highfrequency estimation over a short time interval, whereas Visual-Inertial Odometry, through the use of lower frequencies exteroceptive sensors, remains reliable over a longer period. As both odometries are available on the robot and the experiments do not require a high-frequency state estimation, it was decided to use the Visual-Inertial Odometry instead of the Kinematic-Inertial Odometry, used in the initial system [10], [13].…”

Section: Visual-inertial Odometry Initializationmentioning

confidence: 99%

ICP Localization and Walking Experiments on a TALOS Humanoid Robot

Lasguignes

Maroger

Fallon

et al. 2021

2021 20th International Conference on Advanced Robotics (ICAR)

View full text Add to dashboard Cite

This system paper describes the integration and the evaluation of an ICP-based localization system on the TALOS humanoid robot. The new generation of flash LiDAR systems, here an Ouster OS1-64, have made it possible to obtain 3D clouds at 10 Hz. Coupled with an Intel RealSense T265 providing visual-inertial odometry it is possible to localize the robot and use this information to generate foot steps in real time to reach specific points. The approach is validated with a Qualisys motion capture system. It is also used to generate real-time walking motion on the TALOS humanoid robot. This paper is an integration paper showing that it is now feasible to accurately guide a humanoid robot in an environment in real time using a LiDAR system.

show abstract

Section: Visual-inertial Odometry Initializationmentioning

confidence: 99%

ICP Localization and Walking Experiments on a TALOS Humanoid Robot

Lasguignes

Maroger

Fallon

et al. 2021

2021 20th International Conference on Advanced Robotics (ICAR)

View full text Add to dashboard Cite

show abstract

“…SLAM describes the problem of mapping the environment while self-localizing in the map being built. In order to accomplish these tasks, multiple sensors can be used, including Lidar [ 7 ], inertial measurement units (IMUs) [ 8 ], cameras (RGB [ 6 ], stereo [ 9 ], RGB-D [ 10 ]) or a combination of visual and inertial sensors [ 11 ]. SLAM that only uses cameras as sensors, is usually called visual SLAM, referred to as vSLAM in this paper, and is the one discussed in this work.…”

Section: State-of-the-artmentioning

confidence: 99%

Direct and Indirect vSLAM Fusion for Augmented Reality

2021

View full text Add to dashboard Cite

Augmented reality (AR) is an emerging technology that is applied in many fields. One of the limitations that still prevents AR to be even more widely used relates to the accessibility of devices. Indeed, the devices currently used are usually high end, expensive glasses or mobile devices. vSLAM (visual simultaneous localization and mapping) algorithms circumvent this problem by requiring relatively cheap cameras for AR. vSLAM algorithms can be classified as direct or indirect methods based on the type of data used. Each class of algorithms works optimally on a type of scene (e.g., textured or untextured) but unfortunately with little overlap. In this work, a method is proposed to fuse a direct and an indirect methods in order to have a higher robustness and to offer the possibility for AR to move seamlessly between different types of scenes. Our method is tested on three datasets against state-of-the-art direct (LSD-SLAM), semi-direct (LCSD) and indirect (ORBSLAM2) algorithms in two different scenarios: a trajectory planning and an AR scenario where a virtual object is displayed on top of the video feed; furthermore, a similar method (LCSD SLAM) is also compared to our proposal. Results show that our fusion algorithm is generally as efficient as the best algorithm both in terms of trajectory (mean errors with respect to ground truth trajectory measurements) as well as in terms of quality of the augmentation (robustness and stability). In short, we can propose a fusion algorithm that, in our tests, takes the best of both the direct and indirect methods.

show abstract

“…For instance, in Lidar Odometry And Mapping (LOAM), lidar data are represented as point clouds and robot relative location is achieved by directly matching edge and planar features extracted from point clouds [3]. Recent advances in machine learning algorithms have further accelerated this trend [4,5].…”

Section: Introductionmentioning

confidence: 99%

A physics perspective on lidar data assimilation for mobile robots

Berquin

Zell

2021

Robotica

View full text Add to dashboard Cite

This paper presents a new algorithm for lidar data assimilation relying on a new forward model. Current mapping algorithms suffer from multiple shortcomings, which can be related to the lack of clear forward model. In order to address these issues, we provide a mathematical framework where we show how the use of coarse model parameters results in a new data assimilation problem. Understanding this new problem proves essential to derive sound inference algorithms. We introduce a model parameter specifically tailored for lidar data assimilation, which closely relates to the local mean free path. Using this new model parameter, we derive its associated forward model and we provide the resulting mapping algorithm. We further discuss how our proposed algorithm relates to usual occupancy grid mapping. Finally, we present an example with real lidar measurements.

show abstract

Online LiDAR-SLAM for Legged Robots with Robust Registration and Deep-Learned Loop Closure

Cited by 33 publications

References 28 publications

ICP Localization and Walking Experiments on a TALOS Humanoid Robot

ICP Localization and Walking Experiments on a TALOS Humanoid Robot

Direct and Indirect vSLAM Fusion for Augmented Reality

A physics perspective on lidar data assimilation for mobile robots

Contact Info

Product

Resources

About