Abstract:Background Image matching is crucial in numerous computer vision tasks such as 3D reconstruction and simultaneous visual localization and mapping. The accuracy of the matching significantly impacted subsequent studies. Because of their local similarity, when image pairs contain comparable patterns but feature pairs are positioned differently, incorrect recognition can occur as global motion consistency is disregarded. Methods This study proposes an image-matching filtering algorithm based on global motion consistency. It can be used as a subsequent matching filter for the initial matching results generated by other matching algorithms based on the principle of motion smoothness. A particular matching algorithm can first be used to perform the initial matching;then, the rotation and movement information of the global feature vectors are combined to effectively identify outlier matches. The principle is that if the matching result is accurate, the feature vectors formed by any matched point should have similar rotation angles and moving distances. Thus, global motion direction and global motion distance consistencies were used to reject outliers caused by similar patterns in different locations. Results Four datasets were used to test the effectiveness of the proposed method. Three datasets with similar patterns in different locations were used to test the results for similar images that could easily be incorrectly matched by other algorithms, and one commonly used dataset was used to test the results for the general image-matching problem. The experimental results suggest that the proposed method is more accurate than other state-of-the-art algorithms in identifying mismatches in the initial matching set. Conclusions The proposed outlier rejection matching method can significantly improve the matching accuracy for similar images with locally similar feature pairs in different locations and can provide more accurate matching results for subsequent computer vision tasks.

Rejecting Outliers Based on Correspondence Manifold

Accurate Range Image Registration: Eliminating or Modelling Outliers

Learning an Image Manifold for Retrieval

Learning Correspondence View with Support Vector Machine

Iterative Outlier Removal Using Close-and-Compatible Neighbors for Robust Feature Correspondence

Robust Outlier Removal Using Penalized Linear Regression in Multiview Geometry

Rejecting Mismatches Between Fish-Eye Camera Images by RVM

Outliers Rejection in Similar Image Matching

A Novel Image Registration Method Based on Geometrical Outlier Removal

Outlier Robust Feature Correspondence by Learning Based Matching Process

Robust Outlier Rejection for 3D Registration with Variational Bayes

Comparative Research on Correspondence Function

A High Precision Feature Matching Method Based On Geometrical Outlier Removal For Remote Sensing Image Registration

Rejecting Mismatches by Correspondence Function

The research of detection of outliers based on manifold learning

A New Outlier Removal Strategy Based on Reliability of Correspondence Graph for Fast Point Cloud Registration

Learning High-Dimensional Correspondence Via Manifold Learning and Local Approximation

Partial correspondence based on subgraph matching

A Self-Representation Method with Local Similarity Preserving for Fast Multi-View Outlier Detection

Robust Outlier Rejection From Optical Flow Tracking Points

Learning Coherent Vector Fields for Robust Point Matching under Manifold Regularization