Correspondence matching in unorganized 3D point clouds using Convolutional Neural Networks

作者:

Highlights:

• Contribution to 3D keypoint matching for unorganized point clouds using CNNs

• 3D neighborhoods projected into 2D RGBD patches for 3D keypoint matching

• Proposed correspondence matching with CNNs outperforms hand-crafted descriptors.

摘要

•Contribution to 3D keypoint matching for unorganized point clouds using CNNs•3D neighborhoods projected into 2D RGBD patches for 3D keypoint matching•Proposed correspondence matching with CNNs outperforms hand-crafted descriptors.

论文关键词:Matching,Point cloud,Convolutional Neural Networks

论文评审过程:Received 20 December 2017, Revised 10 October 2018, Accepted 27 February 2019, Available online 8 March 2019, Version of Record 1 April 2019.

论文官网地址:https://doi.org/10.1016/j.imavis.2019.02.013