Targetless Extrinsic Calibration of Multiple Small FoV LiDARs and Cameras Using Adaptive Voxelization

Liu, Xiyuan; Yuan, Chongjian; Zhang, Fu

doi:10.1109/tim.2022.3176889

Cited by 53 publications

(27 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We demonstrate the advantage of the proposed BA approach in multi-lidar calibration with the latest state-of-theart calibration methods based on ICP [22] and BALM [69]. The ICP-based method [22] optimizes the extrinsic parameter R L k T and base lidar pose G R T j by repeatedly registering the point cloud by each lidar at each scan to the rest lidar points until convergence, whereas in both BALM-based method [69] and our method, the extrinsic parameter R L k T and base lidar pose G R T j are optimized concurrently. In [22] the feature correspondence searching is conducted by a k-d tree data structure, while in [69] and our method, this is resolved by adaptive voxelization proposed in BALM [32].…”

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

“…With the ability of concurrent optimization of multiple lidar poses, our BA method can be applied to multi-lidar extrinsic calibration. We consider the problem in [69], which aims to calibrate the extrinsic of multiple solid-state lidars shown in Fig. 12.…”

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

“…The ICP-based method [22] optimizes the extrinsic parameter R L k T and base lidar pose G R T j by repeatedly registering the point cloud by each lidar at each scan to the rest lidar points until convergence, whereas in both BALM-based method [69] and our method, the extrinsic parameter R L k T and base lidar pose G R T j are optimized concurrently. In [22] the feature correspondence searching is conducted by a k-d tree data structure, while in [69] and our method, this is resolved by adaptive voxelization proposed in BALM [32]. To restrain the computation time, the point number in each voxel of [69] have been down-sampled to four, whereas in our method, all feature points are used.…”

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

“…In [22] the feature correspondence searching is conducted by a k-d tree data structure, while in [69] and our method, this is resolved by adaptive voxelization proposed in BALM [32]. To restrain the computation time, the point number in each voxel of [69] have been down-sampled to four, whereas in our method, all feature points are used.…”

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

“…Since the extrinsic between the internal lidars of MID-100 (i.e., L1 L0 T, L1 L2 T) is known fromt he manufacturer's in-factory calibration, it is thus served as the ground truth in both setups. Moreover, in the MID-100 self calibration setup, the middle L 1 is chosen as the reference lidar to calibrate the extrinsic of adjacent lidars, i.e., L1 L0 T, L1 L2 T. This lidar setup is tested with data collected in both [22] and [69] under two scenes which contributes to [22], BALM-based [69] and our proposed method in two experiment setups (with small or without FoV overlap). eight calibration results.…”

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

See 4 more Smart Citations

Efficient and Consistent Bundle Adjustment on Lidar Point Clouds

Liu¹,

Liu²,

Zhang³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Bundle Adjustment (BA) refers to the problem of simultaneous determination of sensor poses and scene geometry, which is a fundamental problem in robot vision. This paper presents an efficient and consistent bundle adjustment method for lidar sensors. The method employs edge and plane features to represent the scene geometry, and directly minimizes the natural Euclidean distance from each raw point to the respective geometry feature. A nice property of this formulation is that the geometry features can be analytically solved, drastically reducing the dimension of the numerical optimization. To represent and solve the resultant optimization problem more efficiently, this paper then proposes a novel concept point clusters, which encodes all raw points associated to the same feature by a compact set of parameters, the point cluster coordinates. We derive the closed-form derivatives, up to the second order, of the BA optimization based on the point cluster coordinates and show their theoretical properties such as the null spaces and sparsity. Based on these theoretical results, this paper develops an efficient second-order BA solver. Besides estimating the lidar poses, the solver also exploits the second order information to estimate the pose uncertainty caused by measurement noises, leading to consistent estimates of lidar poses. Moreover, thanks to the use of point cluster, the developed solver fundamentally avoids the enumeration of each raw point (which is very time-consuming due to the large number) in all steps of the optimization: cost evaluation, derivatives evaluation and uncertainty evaluation.The proposed method is extensively evaluated at different levels: consistency, accuracy, and computation efficiency in both simulated and actual environments. Benchmark evaluation on 19 real-world open sequences covering various datasets (Hilti, NTU-VIRAL and UrbanLoco), environments (campus, urban streets, offices, laboratory, and construction sites), lidar types (Ouster OS0-64, Ouster OS1-16, Velodyne HDL 32E), and motion types (handheld, UAV-based, and ground vehicles-based) shows that our method achieves consistently and significantly higher performance than other state-of-the-art counterparts in terms of localization accuracy, mapping quality, and computation efficiency. In particular, our method achieves a mapping accuracy at a level of the lidar measurement noise (i.e., a few centimeters) while processing all sequences in less than half minute on a standard desktop CPU. Finally, we show how our proposed method effectively improves the accuracy and/or computation efficiency of some important robotic techniques, including lidarinertial odometry, multi-lidar extrinsic calibration, and highaccuracy global mapping. The implementation of our method is open sourced to benefit the robotics community and beyond 2 .

show abstract

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

Section: B Multiple-lidar Calibrationmentioning

confidence: 99%

See 3 more Smart Citations

Efficient and Consistent Bundle Adjustment on Lidar Point Clouds

Liu¹,

Liu²,

Zhang³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Automatic targetless LiDAR–camera calibration: a survey

Xiao

Wang³

et al. 2022

Artif Intell Rev

View full text Add to dashboard Cite

The recent trend of fusing complementary data from LiDARs and cameras for more accurate perception has made the extrinsic calibration between the two sensors critically important. Indeed, to align the sensors spatially for proper data fusion, the calibration process usually involves estimating the extrinsic parameters between them. Traditional LiDAR-camera calibration methods often depend on explicit targets or human intervention, which can be prohibitively expensive and cumbersome. Recognizing these weaknesses, recent methods usually adopt the autonomic targetless calibration approach, which can be conducted at a much lower cost. This paper presents a thorough review of these automatic targetless LiDAR-camera calibration methods. Specifically, based on how the potential cues in the environment are retrieved and utilized in the calibration process, we divide the methods into four categories: information theory based, feature based, ego-motion based, and learning based methods. For each category, we provide an in-depth overview with insights we have gathered, hoping to serve as a potential guidance for researchers in the related fields.

show abstract

LiDAR odometry survey: recent advancements and remaining challenges

Lee,

Jung,

Yang

et al. 2024

Intel Serv Robotics

View full text Add to dashboard Cite

Odometry is crucial for robot navigation, particularly in situations where global positioning methods like global positioning system are unavailable. The main goal of odometry is to predict the robot’s motion and accurately determine its current location. Various sensors, such as wheel encoder, inertial measurement unit (IMU), camera, radar, and Light Detection and Ranging (LiDAR), are used for odometry in robotics. LiDAR, in particular, has gained attention for its ability to provide rich three-dimensional (3D) data and immunity to light variations. This survey aims to examine advancements in LiDAR odometry thoroughly. We start by exploring LiDAR technology and then scrutinize LiDAR odometry works, categorizing them based on their sensor integration approaches. These approaches include methods relying solely on LiDAR, those combining LiDAR with IMU, strategies involving multiple LiDARs, and methods fusing LiDAR with other sensor modalities. In conclusion, we address existing challenges and outline potential future directions in LiDAR odometry. Additionally, we analyze public datasets and evaluation methods for LiDAR odometry. To our knowledge, this survey is the first comprehensive exploration of LiDAR odometry.

show abstract

Targetless Extrinsic Calibration of Multiple Small FoV LiDARs and Cameras Using Adaptive Voxelization

Cited by 53 publications

References 30 publications

Efficient and Consistent Bundle Adjustment on Lidar Point Clouds

Efficient and Consistent Bundle Adjustment on Lidar Point Clouds

Automatic targetless LiDAR–camera calibration: a survey

LiDAR odometry survey: recent advancements and remaining challenges

Contact Info

Product

Resources

About