Learning-based Natural Geometric Matching with Homography Prior

07/13/2018
by   Yifang Xu, et al.
0

Geometric matching is a key step in computer vision tasks. Previous learning-based methods for geometric matching concentrate more on improving alignment quality, while we argue the importance of naturalness issue simultaneously. To deal with this, firstly, Pearson correlation is applied to handle large intra-class variations of features in feature matching stage. Then, we parametrize homography transformation with 9 parameters in full connected layer of our network, to better characterize large viewpoint variations compared with affine transformation. Furthermore, a novel loss function with Gaussian weights guarantees the model accuracy and efficiency in training procedure. Finally, we provide two choices for different purposes in geometric matching. When compositing homography with affine transformation, the alignment accuracy improves and all lines are preserved, which results in a more natural transformed image. When compositing homography with non-rigid thin-plate-spline transformation, the alignment accuracy further improves. Experimental results on Proposal Flow dataset show that our method outperforms state-of-the-art methods, both in terms of alignment accuracy and naturalness.

READ FULL TEXT

page 2

page 3

page 8

page 9

research
07/18/2017

DCTM: Discrete-Continuous Transformation Matching for Semantic Flow

Techniques for dense semantic correspondence have provided limited abili...
research
03/16/2017

Convolutional neural network architecture for geometric matching

We address the problem of determining correspondences between two images...
research
01/04/2017

Path-following based Point Matching using Similarity Transformation

To address the problem of 3D point matching where the poses of two point...
research
03/31/2021

Convolutional Hough Matching Networks

Despite advances in feature representation, leveraging geometric relatio...
research
11/30/2020

Dual Geometric Graph Network (DG2N) – Zero-Shot Refinement for Dense Shape Correspondence

We provide a novel new approach for aligning geometric models using a du...
research
12/18/2017

Deformable Classifiers

Geometric variations of objects, which do not modify the object class, p...
research
09/18/2018

Multiple Combined Constraints for Image Stitching

Several approaches to image stitching use different constraints to estim...

Please sign up or login with your details

Forgot password? Click here to reset