A Robust method for constructing rotational invariant descriptors
作者：
Highlights：
•
• A method was proposed to encode more than one local feature to each sampling point in the image patch.
• A novel binary local feature operator was proposed, which obtains local feature by comparing the relative intensity information of the cross-selected neighboring points.
• Experimental results show that MIOP and MIROP outperform other evaluated local descriptors.
摘要
•A method was proposed to encode more than one local feature to each sampling point in the image patch.•A novel binary local feature operator was proposed, which obtains local feature by comparing the relative intensity information of the cross-selected neighboring points.•Experimental results show that MIOP and MIROP outperform other evaluated local descriptors.
论文关键词：Multi-neighborhood,Rotation invariance,Reference orientation,Image matching,SIFT
论文评审过程：Received 22 April 2017, Revised 19 October 2017, Accepted 20 October 2017, Available online 31 October 2017, Version of Record 4 November 2017.
论文官网地址：https://doi.org/10.1016/j.image.2017.10.010
