Globally consistent alignment for planar mosaicking via topology analysis

作者：

Highlights：

• A generic framework for mosaicking (ordered and unordered) images captured from an approximately planar scene is proposed.

• Method finding a valid main chain of unordered images efficiently by the Minimum Spanning Tree (MST) algorithm is proposed.

• A global alignment strategy able to keep an optimal balance between global consistency and aligning accuracy is proposed.

摘要

Highlights•A generic framework for mosaicking (ordered and unordered) images captured from an approximately planar scene is proposed.•Method finding a valid main chain of unordered images efficiently by the Minimum Spanning Tree (MST) algorithm is proposed.•A global alignment strategy able to keep an optimal balance between global consistency and aligning accuracy is proposed.

论文关键词：Topology estimation,Reference image,Graph analysis,Global consistency,Image mosaicking

论文评审过程：Received 17 July 2016, Revised 1 November 2016, Accepted 11 January 2017, Available online 13 January 2017, Version of Record 12 March 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.01.020