Representing local shape geometry from multi-view silhouette perspective: A distinctive and robust binary 3D feature

作者：

Highlights：

•

• We are the first to use the 2D silhouette cue for 3D local shape description.

• A novel 3D local binary descriptor is proposed for shape representation.

• The proposed descriptor is robust to noise, clutter, occlusion and many other nuisances.

摘要

•We are the first to use the 2D silhouette cue for 3D local shape description.•A novel 3D local binary descriptor is proposed for shape representation.•The proposed descriptor is robust to noise, clutter, occlusion and many other nuisances.

论文关键词：Silhouette,Binary descriptor,Multi-view,Feature matching,Local reference frame

论文评审过程：Received 17 October 2017, Revised 27 January 2018, Accepted 23 March 2018, Available online 28 March 2018, Version of Record 4 April 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.03.015