computer-vision

Image to Point Cloud Registration

3260 papers • 126 benchmarks • 313 datasets

Given a query image and a scene of point cloud, get the camera pose according to them.

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in image-to-point-cloud-registration

Trend

Dataset

Best Model

Actions

KITTI

Libraries

i

Use these libraries to find image-to-point-cloud-registration models and implementations

Datasets

KITTI

Subtasks

No subtasks available.

Most implemented papers

Towards 3D Path Planning from a Single 2D Fluoroscopic Image for Robot Assisted Fenestrated Endovascular Aortic Repair

Jian-Qing Zheng, Xiao-Yun Zhou, C. Riga, Guang-Zhong Yang•Sat Sep 15 2018

A skeleton instantiation framework of Abdominal Aortic Aneurysm (AAA) from a single 2D fluoroscopic image is introduced for real-time 3D robotic path planning and performance advantages were achieved in terms of accuracy, robustness and time-efficiency.

25

Content

0

Paper Graph

Learning General and Distinctive 3D Local Deep Descriptors for Point Cloud Registration

Davide Boscaini, Fabio Poiesi•Thu May 20 2021

These descriptors outperform most recent descriptors by a large margin in terms of generalisation, and also become the state of the art in benchmarks where training and testing are performed in the same domain.

97 0

Paper Graph

CorrI2P: Deep Image-to-Point Cloud Registration via Dense Correspondence

Junhui Hou, Siyu Ren, Yiming Zeng, Xiaodong Chen•Mon Jul 11 2022

The first feature-based dense correspondence framework for addressing the challenging problem of 2D image-to-3D point cloud registration, dubbed CorrI2P is proposed, which outperforms state-of-the-art image- to-point cloud registration methods significantly.

93 0

Paper Graph

GeoTransformer: Fast and Robust Point Cloud Registration With Geometric Transformer

Slobodan Ilic, Zheng Qin, Hao Yu, Changjian Wang, Yulan Guo, Yuxing Peng, Dewen Hu, Kai Xu•Sun Mar 19 2023

Geometric Transformer is proposed, or GeoTransformer for short, to learn geometric feature for robust superpoint matching, making it invariant to rigid transformation and robust in low-overlap cases.

188 0

Paper Graph

FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators

Haiping Wang, Yuan Liu, Bing Wang, Yujing Sun, Zhenchao Dong, Wenping Wang, Bisheng Yang•Wed Oct 04 2023

This work shows that the intermediate features, called diffusion features, extracted by depth-to-image diffusion models are semantically consistent between images and point clouds, which enables the building of coarse but robust cross-modality correspondences.

27 0

Paper Graph

Addressing the generalization of 3D registration methods with a featureless baseline and an unbiased benchmark

T. Pribanić, David Bojanić, Kristijan Bartol, Josep Forest, Tomislav Petkovic•Fri Mar 22 2024

A simple featureless traditional 3D registration baseline method based on the weighted cross-correlation between two given point clouds, which achieves strong results on current benchmarking datasets, outperforming most deep learning methods.

2 0

Paper Graph

Colmap-PCD: An Open-source Tool for Fine Image-to-point cloud Registration

Chunge Bai, Ruiji Fu, Xiang Gao•Sun Oct 08 2023

To the knowledge, this method is the first to register images onto the point cloud map without requiring synchronous capture of camera and LiDAR data, granting the flexibility to manage reconstruction detail levels across various areas of interest.

13 0

Paper Graph

EP2P-Loc: End-to-End 3D Point to 2D Pixel Localization for Large-Scale Visual Localization

Minjung Kim, Junseo Koo, Gunhee Kim•Wed Sep 13 2023

EP2P-Loc is proposed, a novel large-scale visual localization method that mitigates such appearance discrepancy and enables end-to-end training for pose estimation and achieves the state-of-the-art performance compared to existing visual localization and image- to-point cloud registration methods.

20 0

Paper Graph

DeepI2P: Image-to-Point Cloud Registration via Deep Classification

Jiaxin Li, G. Lee•Wed Apr 07 2021

DeepI2P is presented, a novel approach for cross-modality registration between an image and a point cloud that estimates the relative rigid transformation between the co-ordinate frames of the camera and Lidar.

111 0

Paper Graph

Adding a benchmark result helps the community track progress.