Abstract:We present DeepICP - a novel end-to-end learning-based 3D point cloud registration framework that achieves comparable registration accuracy to prior state-of-the-art geometric methods. Different from other keypoint based methods where a RANSAC procedure is usually needed, we implement the use of various deep neural network structures to establish an end-to-end trainable network. Our keypoint detector is trained through this end-to-end structure and enables the system to avoid the inference of dynamic objects, leverages the help of sufficiently salient features on stationary objects, and as a result, achieves high robustness. Rather than searching the corresponding points among existing points, the key contribution is that we innovatively generate them based on learned matching probabilities among a group of candidates, which can boost the registration accuracy. Our loss function incorporates both the local similarity and the global geometric constraints to ensure all above network designs can converge towards the right direction. We comprehensively validate the effectiveness of our approach using both the KITTI dataset and the Apollo-SouthBay dataset. Results demonstrate that our method achieves comparable or better performance than the state-of-the-art geometry-based methods. Detailed ablation and visualization analysis are included to further illustrate the behavior and insights of our network. The low registration error and high robustness of our method makes it attractive for substantial applications relying on the point cloud registration task.

Deep Bayesian ICP Covariance Estimation

Probabilistic 3D ICP algorithm based on ORB feature

A New Approach to 3D ICP Covariance Estimation

Estimating Motion Uncertainty with Bayesian ICP

Em-Based Point to Plane Icp for 3d Simultaneous Localization and Mapping

Stein ICP for Uncertainty Estimation in Point Cloud Matching

BO-ICP: Initialization of Iterative Closest Point Based on Bayesian Optimization

On the Covariance of ICP-based Scan-matching Techniques

DeepICP: An End-to-End Deep Neural Network for 3D Point Cloud Registration

Simulation of Registration Accuracy of Iterative Closest Point (ICP) Method for Pose Estimation

CoBigICP: Robust and Precise Point Set Registration using Correntropy Metrics and Bidirectional Correspondence

An approach to robust ICP initialization

On Designing Consistent Covariance Recovery from a Deep Learning Visual Odometry Engine

Comparing ICP variants on real-world data sets

Geometry-Aware ICP for Scene Reconstruction from RGB-D Camera

DAC: Detector-Agnostic Spatial Covariances for Deep Local Features

DICP: Doppler Iterative Closest Point Algorithm

MAD-ICP: It Is All About Matching Data -- Robust and Informed LiDAR Odometry

TIC-TAC: A Framework for Improved Covariance Estimation in Deep Heteroscedastic Regression

Visual-Inertial Odometry-enhanced Geometrically Stable ICP for Mapping Applications using Aerial Robots

Under Pressure: Altimeter-Aided ICP for 3D Maps Consistency