Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration

Abstract

The Iterative Closest Point (ICP) algorithm is one of the most widely used methods for point-set registration. However, being based on local iterative optimization, ICP is known to be susceptible to local minima. Its performance critically relies on the quality of the initialization and only local optimality is guaranteed. This paper presents the first globally optimal algorithm, named Go-ICP, for Euclidean (rigid) registration of two 3D point-sets under the L2 error metric defined in ICP. The Go-ICP method is based on a branch-and-bound (BnB) scheme that searches the entire 3D motion space SE(3). By exploiting the special structure of SE(3) geometry, we derive novel upper and lower bounds for the registration error function. Local ICP is integrated into the BnB scheme, which speeds up the new method while guaranteeing global optimality. We also discuss extensions, addressing the issue of outlier robustness. The evaluation demonstrates that the proposed method is able to produce reliable registration results regardless of the initialization. Go-ICP can be applied in scenarios where an optimal solution is desirable or where a good initialization is not always available.

to appear in T-PAMI 2016 (IEEE Transactions on Pattern Analysis and Machine Intelligence)

References in corpus (2)

• Point-Set Registration: Coherent Point Drift
• An Adaptive Data Representation for Robust Point-Set Registration and Merging

Cited by in corpus (112)

• RPM-Net: Robust Point Matching using Learned Features
• Vision-based Robotic Grasping From Object Localization, Object Pose Estimation to Grasp Estimation for Parallel Grippers: A Review
• Fast and Robust Iterative Closest Point
• DeepICP: An End-to-End Deep Neural Network for 3D Point Cloud Registration
• DeepIM: Deep Iterative Matching for 6D Pose Estimation
• A comprehensive survey on point cloud registration
• Deep Closest Point: Learning Representations for Point Cloud Registration
• LocNet: Global localization in 3D point clouds for mobile vehicles
• AgriColMap: Aerial-Ground Collaborative 3D Mapping for Precision Farming
• K-means clustering for efficient and robust registration of multi-view point sets
• Canny-VO: Visual Odometry with RGB-D Cameras based on Geometric 3D-2D Edge Alignment
• R-PointHop: A Green, Accurate, and Unsupervised Point Cloud Registration Method
• VRNet: Learning the Rectified Virtual Corresponding Points for 3D Point Cloud Registration
• Localization in Unstructured Environments: Towards Autonomous Robots in Forests with Delaunay Triangulation
• PCRNet: Point Cloud Registration Network using PointNet Encoding
• Automatic Registration between Cone-Beam CT and Scanned Surface via Deep-Pose Regression Neural Networks and Clustered Similarities
• Robust Point Cloud Registration Framework Based on Deep Graph Matching(TPAMI Version)
• Tunnel Facility-based Vehicle Localization in Highway Tunnel using 3D LIDAR
• Globally-Optimal Contrast Maximisation for Event Cameras
• TEASER: Fast and Certifiable Point Cloud Registration
• 3D Point Cloud Processing and Learning for Autonomous Driving
• PHASER: a Robust and Correspondence-free Global Pointcloud Registration
• DPDist : Comparing Point Clouds Using Deep Point Cloud Distance
• Learning and Matching Multi-View Descriptors for Registration of Point Clouds
• Registration of multi-view point sets under the perspective of expectation-maximization
• 3D Point Cloud Registration with Multi-Scale Architecture and Unsupervised Transfer Learning
• Fast Symbolic 3D Registration Solution
• A Sparse and Locally Coherent Morphable Face Model for Dense Semantic Correspondence Across Heterogeneous 3D Faces
• Globally optimal vertical direction estimation in Atlanta World
• Point Cloud Registration using Representative Overlapping Points
• Discriminative Optimization: Theory and Applications to Computer Vision Problems
• RANSAC Back to SOTA: A Two-stage Consensus Filtering for Real-time 3D Registration
• Efficient and Robust Registration on the 3D Special Euclidean Group
• AlignNet-3D: Fast Point Cloud Registration of Partially Observed Objects
• Multiview Point Cloud Registration Based on Minimum Potential Energy for Free-Form Blade Measurement
• Indoor simultaneous localization and mapping based on fringe projection profilometry
• DeepGMR: Learning Latent Gaussian Mixture Models for Registration
• Automatic Robot Hand-Eye Calibration Enabled by Learning-Based 3D Vision
• LPRnet: A self-supervised registration network for LiDAR and photogrammetric point clouds
• A Review on Geometry and Surface Inspection in 3D Concrete Printing
• MULLS: Versatile LiDAR SLAM via Multi-metric Linear Least Square
• Certifiably Correct Range-Aided SLAM
• Intraoperative Liver Surface Completion with Graph Convolutional VAE
• CLIPPER+: A Fast Maximal Clique Algorithm for Robust Global Registration
• Robotic Depowdering for Additive Manufacturing Via Pose Tracking
• Iterative Distance-Aware Similarity Matrix Convolution with Mutual-Supervised Point Elimination for Efficient Point Cloud Registration
• A Polynomial-time Solution for Robust Registration with Extreme Outlier Rates
• One Framework to Register Them All: PointNet Encoding for Point Cloud Alignment
• Fast and Globally Optimal Rigid Registration of 3D Point Sets by Transformation Decomposition
• Learning 3D-3D Correspondences for One-shot Partial-to-partial Registration
• Homomorphic Sensing
• Automatic 3D Registration of Dental CBCT and Face Scan Data using 2D Projection Images
• Polynomial-time algorithms for continuous metrics on atomic clouds of unordered points
• LabelFusion: A Pipeline for Generating Ground Truth Labels for Real RGBD Data of Cluttered Scenes
• DBDNet:Partial-to-Partial Point Cloud Registration with Dual Branches Decoupling
• Shape Registration with Directional Data
• Globally-Optimal Inlier Set Maximisation for Simultaneous Camera Pose and Feature Correspondence
• GMCR: Graph-based Maximum Consensus Estimation for Point Cloud Registration
• An approach to robust ICP initialization
• Accelerating Globally Optimal Consensus Maximization in Geometric Vision
• DeepMapping: Unsupervised Map Estimation From Multiple Point Clouds
• Learning the shape of female breasts: an open-access 3D statistical shape model of the female breast built from 110 breast scans
• Unsupervised 3D Learning for Shape Analysis via Multiresolution Instance Discrimination
• A Novel Method for the Absolute Pose Problem with Pairwise Constraints
• SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondences
• Elastic Registration of Geodesic Vascular Graphs
• Fast Rigid 3D Registration Solution: A Simple Method Free of SVD and Eigen-Decomposition
• Hybrid Trilinear and Bilinear Programming for Aligning Partially Overlapping Point Sets
• ROBIN: a Graph-Theoretic Approach to Reject Outliers in Robust Estimation using Invariants
• DeepVIO: Self-supervised Deep Learning of Monocular Visual Inertial Odometry using 3D Geometric Constraints
• Rotation-Invariant Rapid TRISO-Fueled Pebble Identification Based on Feature Matching and Point Cloud Registration
• CLIPPER: Robust Data Association without an Initial Guess
• One Ring to Rule Them All: Certifiably Robust Geometric Perception with Outliers
• Equi-GSPR: Equivariant SE(3) Graph Network Model for Sparse Point Cloud Registration
• Multiview Point Cloud Registration via Optimization in an Autoencoder Latent Space
• Deep-3DAligner: Unsupervised 3D Point Set Registration Network With Optimizable Latent Vector
• 3DMNDT:3D multi-view registration method based on the normal distributions transform
• DFC: Deep Feature Consistency for Robust Point Cloud Registration
• Efficient and Deterministic Search Strategy Based on Residual Projections for Point Cloud Registration with Correspondences
• Robust Multi-view Registration of Point Sets with Laplacian Mixture Model
• Point-Pattern Matching Technique for Local Structural Analysis in Condensed Matter
• From noisy point clouds to complete ear shapes: unsupervised pipeline
• Robust Multi-Robot Global Localization with Unknown Initial Pose based on Neighbor Constraints
• Automatic clustering of Celtic coins based on 3D point cloud pattern analysis
• Enriching thermal point clouds of buildings using semantic 3D building models
• Target-less registration of point clouds: A review
• Partial Computing Offloading Assisted Cloud Point Registration in Multi-robot SLAM
• Regularized Wasserstein Means for Aligning Distributional Data
• A Direct Semi-Exhaustive Search Method for Robust, Partial-to-Full Point Cloud Registration
• Can generalised relative pose estimation solve sparse 3D registration?
• Adaptive Cost Function for Pointcloud Registration
• Globally optimal consensus maximization for robust visual inertial localization in point and line map
• Review on 6D Object Pose Estimation with the focus on Indoor Scene Understanding
• Point Cloud Registration Based on Consistency Evaluation of Rigid Transformation in Parameter Space
• A Method to Generate High Precision Mesh Model and RGB-D Datasetfor 6D Pose Estimation Task
• Adversarial point set registration
• Multi-view registration of unordered range scans by fast correspondence propagation of multi-scale descriptors
• What Stops Learning-based 3D Registration from Working in the Real World?
• 3D Unsupervised Region-Aware Registration Transformer
• Structured Analytic Mappings for Point Set Registration
• Loop Closure via Maximal Cliques in 3D LiDAR-Based SLAM
• Rethinking PointNet Embedding for Faster and Compact Model
• A Robust Loss for Point Cloud Registration
• PRANet: Point Cloud Registration with an Artificial Agent
• Towards Fine-grained 3D Face Dense Registration: An Optimal Dividing and Diffusing Method
• 3D Registration for Self-Occluded Objects in Context
• Effective multi-view registration of point sets based on student's t mixture model
• A Second-Order Lower Bound for Globally Optimal 2D Registration
• Performance Analysis of NDT-based Graph SLAM for Autonomous Vehicle in Diverse Typical Driving Scenarios of Hong Kong
• Quasi Branch and Bound for Smooth Global Optimization
• Recurrent Convolutional Neural Networks for LiDAR-Based Attitude Initialization of Rotating Spacecraft
• Least-squares registration of point sets over SE (d) using closed-form projections