Keypoint descriptor matching with context-based orientation estimation
作者：
Highlights：
• Novel matching strategies for histogram-based descriptors are presented.
• Global dominant orientation is used by exploiting the image context.
• A new 3D extensible framework to evaluate feature descriptors is introduced.
• 2D/3D comparisons with state-of-the-art rotational invariant descriptors are reported.
• Results show the effectiveness of the proposed matching approaches.
摘要
•Novel matching strategies for histogram-based descriptors are presented.•Global dominant orientation is used by exploiting the image context.•A new 3D extensible framework to evaluate feature descriptors is introduced.•2D/3D comparisons with state-of-the-art rotational invariant descriptors are reported.•Results show the effectiveness of the proposed matching approaches.
论文关键词：Image descriptors,Local features,Dominant orientation,Rotation invariance,Keypoint matching,SIFT,LIOP,MROGH
论文评审过程：Received 9 December 2013, Revised 17 April 2014, Accepted 22 May 2014, Available online 1 June 2014.
论文官网地址：https://doi.org/10.1016/j.imavis.2014.05.002
