Scale-invariant Feature Descriptor under Triangular Combination Constraint

Zhiheng Wang,Hongya Zhang,Hongmin Liu,Jing Wang
DOI: https://doi.org/10.3969/j.issn.1003-9775.2017.02.005
2017-01-01
Abstract:In this paper,we propose a scale-invariant feature descriptor under triangular combination constraint to overcome the issues of less number of the matched points and low matching accuracy for point descriptor.First,three feature points under given geometric constraints are combined into a triangle.Then,using the radius of the inscribed circle of the obtained triangle as the support region size,a scale-invariant descriptor for the triangle is thus created and the matching is made subsequently.Finally,the point correspondences are obtained by the computed main direction of the support region,and mismatches are eliminated by retaining the most repeated point match.The experimental results show that this method is not only robust to various image transforms,such as rotation,viewpoint change and JPEG compression,but also increases the number of the matched points with high accuracy.
What problem does this paper attempt to address?