Abstract:For the problems of large viewpoint variation, heavy distortion and small overlap area in UAV images registration, this paper proposes a density analysis based method to remove mismatches in putative feature correspondences. Our method uses intra-cluster topological constraints for mismatch filtering, which is based on a density-based hierarchical clustering algorithm. Compared with other methods that perform mismatch filtering based on neighborhood topological relationship, our method is more robust to viewpoint changes both in horizontal and vertical directions. The algorithm in this paper uses a coarse-to-fine strategy, which starts with establishing putative feature correspondences based on local descriptors, such as SIFT, ORB, etc. After that we focus on removing outliers by clustering these feature points and verifying topology consistency of the clusters in different images. We view the feature point matching problem as a correspondence problem of the same visual model in two images, and clustering the feature points based on density can approximate the separation of multiple visual patterns. We tested our algorithm on a UAV image dataset which includes several pairs of images and their ground truth. These image pairs contain viewpoint changes in horizontal, vertical and their mixture which produce problems of low overlap, image distortion and severe outliers. Experiments demonstrate that our method significantly outperforms the state-of-the-art in terms of matching precision.

Discarding wide baseline mismatches via topological clustering

Topological Clustering and Its Application for Discarding Wide-Baseline Mismatches

Topologically Clustering: a Method for Discarding Mismatches

Hierarchical Clustering Matching for Features with Repetitive Patterns in Visual Odometry

A Matching Algorithm Based On Correlations Between Matrices Of Reference Differences Combining Disparities

Image Matching Optimization Based on Taguchi Method and Adaptive Spatial Clustering with SIFT Features.

Machine Learning on Topological Constraint for Mismatching Removal.

Removal of Mismatches in Scale-Invariant Feature Transform Algorithm Using Image Depth Information

Local Topology Similarity Guided Probabilistic Sampling for Mismatch Removal

Adaptive clustering feature matching algorithm based on SIFT and RANSAC

A Reliable Algorithm for Image Matching Based on SIFT

Mismatches Removal in Unmanned Aerial Vehicle Image Registration Using Density Based Method

Robust Feature Matching Using Spatial Clustering With Heavy Outliers

Adaptively feature matching via joint transformational-spatial clustering

Wide Baseline Stereo Matching Based on SIFT and SVD

Large disparity motion layer extraction via topological clustering.

A Two Stage Iterative Procrustes Matching Algorithm Based on SIFT Feature

Mismatching Removal for Feature-Point Matching Based on Triangular Topology Probability Sampling Consensus

LTM: Efficient Learning with Triangular Topology Constraint for Feature Matching with Heavy Outliers

Contourlet-SIFT Feature Matching Algorithm

Robust Image Matching Algorithm Using Sift on Multiple Layered Strategies