Point cloud registration means aligning pairs of point clouds that lie in different positions and orientations, which contains global registration and local registration. Global registration means coarse registration, where the pairs of point clouds have large transformations, and global registration provides an initial alignment. Local registration means fine registration, where the poses of the point clouds have little difference.
2020:
[arXiv] When Deep Learning Meets Data Alignment: A Review on Deep Registration Networks (DRNs), [paper]
[arXiv] Least Squares Optimization: from Theory to Practice, [paper]
2012:
[TVCG] Registration of 3D Point Clouds and Meshes: A Survey From Rigid to Non-Rigid, [[paper](http://orca.cf.ac.uk/47333/1/ROSIN registration of 3d point clouds and meshes.pdf)]
Most of the global registration methods operate on candidate correspondences. Other approaches are based on the branch-and-bound techniques, which explore the pose space exhaustively.
2020:
[ECCV] DH3D: Deep Hierarchical 3D Descriptors for Robust Large-Scale 6DoF Relocalization, [paper]
[ECCV] Iterative Distance-Aware Similarity Matrix Convolution with Mutual-Supervised Point Elimination for Efficient Point Cloud Registration, [paper] [code]
[PRL] Fuzzy Logic and Histogram of Normal Orientation-based 3D Keypoint Detection for Point Clouds, [paper]
[arXiv] Learning 3D-3D Correspondences for One-shot Partial-to-partial Registration, [paper]
[arXiv] RPM-Net: Robust Point Matching using Learned Features, [paper]
[arXiv] End-to-End Learning Local Multi-view Descriptors for 3D Point Clouds, [paper]
[arXiv] D3Feat: Joint Learning of Dense Detection and Description of 3D Local Features, [paper]
[arXiv] Self-supervised Point Set Local Descriptors for Point Cloud Registration, [paper]
[arXiv] StickyPillars: Robust feature matching on point clouds using Graph Neural Networks, [paper]
[arXiv] LRF-Net: Learning Local Reference Frames for 3D Local Shape Description and Matching, [paper]
2019:
[CVPR] 3DRegNet: A Deep Neural Network for 3D Point Registration, [paper] [code]
[ICCV] DeepICP: An End-to-End Deep Neural Network for 3D Point Cloud Registration, [paper]
[ICCV] Deep Closest Point: Learning Representations for Point Cloud Registration, [paper] [code]
[CVPR] The Perfect Match: 3D Point Cloud Matching with Smoothed Densities, [paper]
2018:
[ECCV] 3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration, [paper] [code]
2017:
[CVPR] 3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions, [paper] [code]
2016:
[ECCV] Fast Global Registration, [paper] [code]
[IJCV] A comprehensive performance evaluation of 3D local feature descriptors, [paper]
2014:
[CVIU] SHOT: Unique signatures of histograms for surface and texture description, [paper)]
[SGP] Super 4PCS: Fast Global Pointcloud Registration via Smart Indexing, [paper] [code]
2012:
[IJRR] Rigid 3D Geometry Matching for Grasping of Known Objects in Cluttered Scenes, [paper] [code]
2011:
[ICCVW] CAD-model recognition and 6DOF pose estimation using 3D cues, [paper]
2010:
[CVPR] Model Globally, Match Locally: Efficient and Robust 3D Object Recognition, [paper] [code]
2009:
[ICRA] Fast Point Feature Histograms (FPFH) for 3D registration, [paper]
2008:
[TOG] 4-points congruent sets for robust pairwise surface registration, [paper]
1981:
[ACM] Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography, [paper]
2020:
[arXiv] DeepGMR: Learning Latent Gaussian Mixture Models for Registration, [paper]
[ITSC] DeepCLR: Correspondence-Less Architecture for Deep End-to-End Point Cloud Registration, [paper]
[arXiv] Aligning Partially Overlapping Point Sets: an Inner Approximation Algorithm, [paper]
[arXiv] Minimum Potential Energy of Point Cloud for Robust Global Registration, [paper]
[arXiv] A Dynamical Perspective on Point Cloud Registration, [paper]
[arXiv] Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondences, [paper]
[CVPR] Deep Global Registration, [paper]
[arXiv] DPDist : Comparing Point Clouds Using Deep Point Cloud Distance, [paper]
[arXiv] Single Shot 6D Object Pose Estimation, [paper]
[arXiv] A Benchmark for Point Clouds Registration Algorithms, [paper] [code]
[arXiv] TEASER: Fast and Certifiable Point Cloud Registration, [paper] [code]
[arXiv] Plane Pair Matching for Efficient 3D View Registration, [paper]
[arXiv] Learning multiview 3D point cloud registration, [paper]
[arXiv] Robust, Occlusion-aware Pose Estimation for Objects Grasped by Adaptive Hands, [paper]
[arXiv] Non-iterative One-step Solution for Point Set Registration Problem on Pose Estimation without Correspondence, [paper]
[arXiv] 6D Object Pose Regression via Supervised Learning on Point Clouds, [paper]
2019:
[arXiv] One Framework to Register Them All: PointNet Encoding for Point Cloud Alignment, [paper]
[arXiv] PCRNet: Point Cloud Registration Network using PointNet Encoding, [paper] [code]
[CVPR] The Alignment of the Spheres: Globally-Optimal Spherical Mixture Alignment for Camera Pose Estimation, [paper]
2016:
[CVPR] GOGMA: Globally-Optimal Gaussian Mixture Alignment, [paper]
[TPAMI] Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration, [paper] [code]
2020:
[arXiv] Unsupervised Learning of 3D Point Set Registration, [paper]
[arXiv] Fast and Robust Iterative Closet Point, [paper]
[arXiv] Applying Lie Groups Approaches for Rigid Registration of Point Clouds, [paper]
[arXiv] An Analysis of SVD for Deep Rotation Estimation, [paper]
2019:
[CVPR] PointNetLK: Robust & Efficient Point Cloud Registration using PointNet, [paper] [code]
[NeurIPS] PRNet: Self-Supervised Learning for Partial-to-Partial Registration, [paper]
[TOG] A symmetric objective function for ICP, [paper]
2009:
[RSS] Generalized ICP, [paper]
2005:
[IVC] Robust Euclidean alignment of 3D point sets: the Trimmed Iterative Closest Point algorithm, [paper]
2004:
[Report] Linear least-squares optimization for point-to-plane icp surface registration, [paper]
2003:
[IVC] Robust registration of 2D and 3D point sets, [paper]
2001:
[3DDIM] Efficient variants of the ICP algorithm, [paper]
1992:
[TPAMI] ICP: A method for registration of 3-D shapes, [paper]
1991:
[ICRA] Object modeling by registration of multiple range images, [paper]
2020:
[arXiv] PointGMM: a Neural GMM Network for Point Clouds, [paper]
2019:
[CVPR] FilterReg: Robust and Efficient Probabilistic Point-Set Registration using Gaussian Filter and Twist Parameterization, [paper] [code]
2018:
[ECCV] HGMR: Hierarchical Gaussian Mixtures for Adaptive 3D Registration, [paper]
[CVPR] Density Adaptive Point Set Registration, [paper]
[CVPR] Fast Monte-Carlo Localization on Aerial Vehicles using Approximate Continuous Belief Representations, [paper]
[RAL] On-Manifold GMM Registration, [paper]
2017:
[TPAMI] Joint Alignment of Multiple Point Sets with Batch and Incremental Expectation-Maximization, [paper]
2012:
[IJRR] Fast and accurate scan registration through minimization of the distance between compact 3D NDT representations, [paper]
2011:
[TPAMI] Robust Point Set Registration Using Gaussian Mixture Models, [paper] [code]
2009:
[D] The three-dimensional normal-distributions transform: an efficient representation for registration, surface analysis, and loop detection, [paper]
2002:
[ECCV] Multi-scale EM-ICP: A Fast and Robust Approach for Surface Registration, [paper]
2020:
[arXiv] SceneCAD: Predicting Object Alignments and Layouts in RGB-D Scans, [paper]
2019:
[ICCV] End-to-End CAD Model Retrieval and 9DoF Alignment in 3D Scans, [paper]
2016:
[arXiv] Lessons from the Amazon Picking Challenge, [paper]
[arXiv] Team Delft's Robot Winner of the Amazon Picking Challenge 2016, [paper]