Globally Consistent 3D LiDAR Mapping With GPU-Accelerated GICP Matching Cost Factors

Koide, Kenji; Yokozuka, Masashi; Oishi, Shuji; Banno, Atsuhiko

doi:10.1109/lra.2021.3113043

Cited by 32 publications

(25 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As pose graph optimization uses SE3 relative pose constraints to constrain sensor poses, separation of the LiDAR-based estimation and IMU-based estimation is unavoidable. It also affects the consistency of the map when closing a large loop or constraining small overlapping frames because it employs an approximated representation (i.e., Gaussian distribution) for relative pose constraints [4].…”

Section: B Lidar-imu Backendmentioning

confidence: 99%

“…2, we obtain a Hessian factor to constrain the relative pose between T i and T j . It is worth emphasizing that we re-evaluate and linearize e M at the current linearization point for every optimization iteration, which results in a more accurate constraint than the traditional SE3 relative pose constraint [4].…”

Section: A Lidar Matching Cost Factormentioning

confidence: 99%

“…To manage keyframes, we define an overlap rate between two frames P i and P j as the fraction of points in P i that fall within a voxel of P j [4]. Every time a new frame arrives, we evaluate the overlap rate between that frame and the latest keyframe and, if the overlap is smaller than a threshold (e.g., 90%), we insert that frame into the keyframe list.…”

Section: Odometry Estimationmentioning

confidence: 99%

“…We use the voxel-based GICP matching cost factor, which can fully leverage GPU parallel processing and enables the creation of a factor graph to minimize the This work was supported in part by a project commissioned by the New Energy and Industrial Technology Development Organization (NEDO). 1 All the authors are with the Department of Information Technology and Human Factors, the National Institute of Advanced Industrial Science and Technology, Umezono 1-1-1, Tsukuba, 3050061, Ibaraki, Japan, k.koide@aist.go.jp scan matching error over the entire map [4]. We combine the GPU-accelerated matching cost factor with the IMU preintegration factor to jointly consider the LiDAR and IMU constraints for global trajectory optimization.…”

Section: Introductionmentioning

confidence: 99%

“…1) It is based on the voxelized GICP matching cost factor [4], which uses a larger number of points to calculate the registration error compared to the commonly used scan matching based on line and plane point matching [1]. This enables accurate and robust constraint of sensor poses while fully leveraging GPU parallel processing.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Globally Consistent and Tightly Coupled 3D LiDAR Inertial Mapping

Koide¹,

Yokozuka²,

Oishi³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

This paper presents a real-time 3D mapping framework based on global matching cost minimization and LiDAR-IMU tight coupling. The proposed framework comprises a preprocessing module and three estimation modules: odometry estimation, local mapping, and global mapping, which are all based on the tight coupling of the GPU-accelerated voxelized GICP matching cost factor and the IMU preintegration factor. The odometry estimation module employs a keyframe-based fixed-lag smoothing approach for efficient and low-drift trajectory estimation, with a bounded computation cost. The global mapping module constructs a factor graph that minimizes the global registration error over the entire map with the support of IMU constraints, ensuring robust optimization in feature-less environments. The evaluation results on the Newer College dataset and KAIST urban dataset show that the proposed framework enables accurate and robust localization and mapping in challenging environments.

show abstract

Section: B Lidar-imu Backendmentioning

confidence: 99%

Section: A Lidar Matching Cost Factormentioning

confidence: 99%

Section: Odometry Estimationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Globally Consistent and Tightly Coupled 3D LiDAR Inertial Mapping

Koide¹,

Yokozuka²,

Oishi³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

A compact RTK‐GNSS device for high‐precision localization of outdoor mobile robots

Lee,

Kim,

Yang

et al. 2024

Journal of Field Robotics

View full text Add to dashboard Cite

Accurate location information is crucial for ensuring the reliability of service robots, such as field vehicles and agricultural robots. These practical robots often encounter fluctuating or inclined terrains in dynamic conditions during operation. Additionally, the performance of mobile robots depends on the physical size and weight of the localization system–physically heavy systems may degrade performance. In this paper, we present a compact, lightweight RTK‐GNSS device weighing under 40 g, which achieves centimeter‐level accuracy across different attitudes and dynamic situations. To enhance signal reception and positioning performance, we lessen noise levels in the power circuit and radio‐frequency signal conditioning using a low‐dropout regulator and low‐noise amplifier. Utilizing the configured device, we verified centimeter‐level accuracy through experiments of tracking an unmanned ground vehicle. The potential of the device is substantiated via three application scenarios—a localization of a moving human for a wearable device, a heading estimation of a mobile robot in stationary conditions, and practical field tests in various signal‐reception environments. Attributable to its compact dimensions and lightweight features, the proposed device can yield a low system strain while maintaining high accuracy and reliable performance, rendering it highly adaptable for various robotic applications.

show abstract

Inquiring the robot operating system community on the state of adoption of the ROS 2 robotics middleware

Portugal,

Rocha,

Castilho

2024

Int J Intell Robot Appl

View full text Add to dashboard Cite

The robot operating system (ROS) is a well-established and popular open-source robotics middleware used for rapid prototyping of robotic applications. However, ROS has several known weaknesses, such as lack of support for real-time systems and limitations in fully distributed deployments with multiple robots. To address these weaknesses, ROS underwent a major update and the first version of ROS 2 was released back in 2015, being lingeringly adopted by the community in recent years. Yet, long-term support for ROS 1 will come to an end on May 2025. As such, it is more important than ever to analyze and explore the features of ROS 2, and understand its technological readiness. This work focuses on the state of adoption of ROS 2. Specifically, the article presents a user experience questionnaire targeting the ROS community to understand the more eminent needs with respect to ROS 2, determine the current levels of adoption, and identify what is holding the community back from migrating their ROS 1 applications to ROS 2. Results with more than 100 participants from around the world show that the community is aware of ROS 2 and the middleware is addressing most user expectations. We unveil the main reasons given for not migrating to ROS 2, further understand the community’s views on ROS 2, and identify what is missing to anticipate ROS 1 users migrating to ROS 2 and make ROS 2 widely adopted. Moreover, we gather important impressions on the appropriateness of ROS 2 for multi-robot system use cases.

show abstract

Globally Consistent 3D LiDAR Mapping With GPU-Accelerated GICP Matching Cost Factors

Cited by 32 publications

References 82 publications

Globally Consistent and Tightly Coupled 3D LiDAR Inertial Mapping

Globally Consistent and Tightly Coupled 3D LiDAR Inertial Mapping

A compact RTK‐GNSS device for high‐precision localization of outdoor mobile robots

Inquiring the robot operating system community on the state of adoption of the ROS 2 robotics middleware

Contact Info

Product

Resources

About