Large-scale point cloud generated from 3D sensors is more accurate than its image-based counterpart. However, it is seldom used in visual pose estimation due to the difficulty in obtaining 2D-3D image to point cloud correspondences. In this paper, we propose the 2D3D-MatchNet - an end-to-end deep network architecture to jointly learn the descriptors for 2D and 3D keypoint from image and point cloud, respectively. As a result, we are able to directly match and establish 2D-3D correspondences from the query image and 3D point cloud reference map for visual pose estimation. We create our Oxford 2D-3D Patches dataset from the Oxford Robotcar dataset with the ground truth camera poses and 2D-3D image to point cloud correspondences for training and testing the deep network. Experimental results verify the feasibility of our approach.
@article{arxiv.1904.09742,
title = {2D3D-MatchNet: Learning to Match Keypoints Across 2D Image and 3D Point Cloud},
author = {Mengdan Feng and Sixing Hu and Marcelo Ang and Gim Hee Lee},
journal= {arXiv preprint arXiv:1904.09742},
year = {2019}
}