3D Point Cloud Processing and Learning for Autonomous Driving: Impacting Map Creation, Localization, and Perception

Chen, Siheng; Liu, Baoan; Feng, Chen; Vallespi-Gonzalez, Carlos; Wellington, Carl

doi:10.1109/msp.2020.2984780

Cited by 140 publications

(48 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Point clouds are important data structures that represent the three-dimensional, real world. Because point clouds have no topological structure and are also easy to store and transmit, they are widely used in 3D reconstruction, autonomous driving, intelligent robots, and many other applications [1][2][3]. However, the point cloud for an object is usually obtained using two or more scans from different reference frames because of the limitation of the geometric shape of the measured object and the scanning angle.…”

Section: Introductionmentioning

confidence: 99%

Automatic Point Cloud Registration for Large Outdoor Scenes Using a Priori Semantic Information

Huang

Cui

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

As an important and fundamental step in 3D reconstruction, point cloud registration aims to find rigid transformation that register two point sets. The major challenge in point cloud registration techniques is finding correct correspondences in the scenes which may contain many repetitive structures and noise. This paper is primarily concerned with improving registration using a priori semantic information in the search for correspondences. In particular, we present a new point cloud registration pipeline for large outdoor scenes that takes advantage of semantic segmentation. Our method consists of extracting semantic segments from point clouds uses an efficient deep neural network; then, detecting the key points of the point cloud and using a feature descriptor to get the initial correspondence set; finally, applying a Random Sample Consensus (RANSAC) strategy to estimate the transformations that align segments with the same labels. Instead of using all points to estimate a global alignment, our method aligns two point clouds using transformations calculated by each segment with the highest inlier ratio. We evaluate our method on the publicly available Whu-TLS registration dataset. These experiments demonstrate how a priori semantic information the improves registration in terms of precision and speed.

show abstract

Section: Introductionmentioning

confidence: 99%

Automatic Point Cloud Registration for Large Outdoor Scenes Using a Priori Semantic Information

Huang

Cui

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The loss function L = αL trans + βL rot + γL cl , where L trans is translation loss, L rot is rotation loss and L cl is contrastive loss. When the model predicts the best transformation R, t, the truth transformation R, t is known, L trans and L rot are defined by Equations ( 9) and ( 10), respectively, and L cl is defined by Equation (3).…”

Section: Loss Functionmentioning

confidence: 99%

“…Point cloud registration is an important and fundamental field in 3D computer vision and graphics. It has many applications, such as 3D reconstruction [1], 3D image fusion [2], simultaneous localization and mapping (SLAM), [3][4][5], among others. In recent years, remarkable progress has been made in the point cloud registration, which aims to align the source to the target point cloud, so as to unify the two into the agreed coordinate system.…”

Section: Introductionmentioning

confidence: 99%

SCRnet: A Spatial Consistency Guided Network Using Contrastive Learning for Point Cloud Registration

et al. 2022

View full text Add to dashboard Cite

Point cloud registration is used to find a rigid transformation from the source point cloud to the target point cloud. The main challenge in the point cloud registration is in finding correct correspondences in complex scenes that may contain many noise and repetitive structures. At present, many existing methods use outlier rejections to help the network obtain more accurate correspondences, but they often ignore the spatial consistency between keypoints. Therefore, to address this issue, we propose a spatial consistency guided network using contrastive learning for point cloud registration (SCRnet), in which its overall stage is symmetrical. SCRnet consists of four blocks, namely feature extraction block, confidence estimation block, contrastive learning block and registration block. Firstly, we use mini-PointNet to extract coarse local and global features. Secondly, we propose confidence estimation block, which formulate outlier rejection as confidence estimation problem of keypoint correspondences. In addition, the local spatial features are encoded into the confidence estimation block, which makes the correspondence possess local spatial consistency. Moreover, we propose contrastive learning block by constructing positive point pairs and hard negative point pairs and using Point-Pair-INfoNCE contrastive loss, which can further remove hard outliers through global spatial consistency. Finally, the proposed registration block selects a set of matching points with high spatial consistency and uses these matching sets to calculate multiple transformations, then the best transformation can be identified by initial alignment and Iterative Closest Point (ICP) algorithm. Extensive experiments are conducted on KITTI and nuScenes dataset, which demonstrate the high accuracy and strong robustness of SCRnet on point cloud registration task.

show abstract

“…P OINT cloud registration is to align two or more 3D point clouds acquired from various views, platforms, or at different times into a unified coordinate system [1]. This technique is the cornerstone of many 3D computer vision applications, such as 3D reconstruction [2], augmenting reality [3], autonomous driving [4], [5], [6], cancer radiotherapy [7], [8] and robotics [9], [10].…”

Section: Introductionmentioning

confidence: 99%

COTReg:Coupled Optimal Transport based Point Cloud Registration

Mei¹,

Huang²,

Yu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Generating a set of high-quality correspondences or matches is one of the most critical steps in point cloud registration. This paper proposes a learning framework COTReg by jointly considering the pointwise and structural matchings to predict correspondences of 3D point cloud registration. Specifically, we transform the two matchings into a Wasserstein distance-based and a Gromov-Wasserstein distance-based optimizations, respectively. Thus the task of establishing the correspondences can be naturally reshaped to a coupled optimal transport problem. Furthermore, we design a network to predict the confidence score of being an inlier for each point of the point clouds, which provides the overlap region information to generate correspondences. Our correspondence prediction pipeline can be easily integrated into either learning-based features like FCGF or traditional descriptors like FPFH. We conducted comprehensive experiments on 3DMatch, KITTI, 3DCSR, and ModelNet40 benchmarks, showing the state-of-art performance of the proposed method.

show abstract

3D Point Cloud Processing and Learning for Autonomous Driving: Impacting Map Creation, Localization, and Perception

Cited by 140 publications

References 26 publications

Automatic Point Cloud Registration for Large Outdoor Scenes Using a Priori Semantic Information

Automatic Point Cloud Registration for Large Outdoor Scenes Using a Priori Semantic Information

SCRnet: A Spatial Consistency Guided Network Using Contrastive Learning for Point Cloud Registration

COTReg:Coupled Optimal Transport based Point Cloud Registration

Contact Info

Product

Resources

About