Fast and Interpretable 2D Homography Decomposition: Similarity-Kernel-Similarity and Affine-Core-Affine Transformations
Not on the map. Why?
Abstract (arXiv)
In this paper, we present two fast and interpretable decomposition methods for 2D homography, which are named Similarity-Kernel-Similarity (SKS) and Affine-Core-Affine (ACA) transformations respectively. Under the minimal 4-point configuration, the first and the last similarity transformations in SKS are computed by two anchor points on target and source planes, respectively. Then, the other two point correspondences can be exploited to compute the middle kernel transformation with only four parameters. Furthermore, ACA uses three anchor points to compute the first and the last affine transformations, followed by computation of the middle core transformation utilizing the other one point correspondence. ACA can compute a homography up to a scale with only 85 floating-point operations (FLOPs), without even any division operations. Therefore, as a plug-in module, ACA facilitates the traditional feature-based Random Sample Consensus (RANSAC) pipeline, as well as deep homography pipelines estimating 4-point offsets. In addition to the advantages of geometric parameterization and computational efficiency, SKS and ACA can express each element of homography by a polynomial of input coordinates (7th degree to 9th degree), extend the existing essential Similarity-Affine-Projective (SAP) decomposition and calculate 2D affine transformations in a unified way. Source codes are released in https://github.com/cscvlab/SKS-Homography.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 22 of its 93 references that are in the map. First posted on arXiv 28 Feb 2024.
Builds on
22 of 93 references.
References · 22
- 2014Microsoft COCO: Common Objects in Contextcited by 17,498
- 2022Instant Neural Graphics Primitives with a Multiresolution Hash Encodingcited by 3,698
- 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,313
- 2020ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAMcited by 862
- 2016LIFT: Learned Invariant Feature Transformcited by 305
- 2016DSAC - Differentiable RANSAC for Camera Localizationcited by 264
- 2020Image Matching across Wide Baselines: From Paper to Practicecited by 224
- 2019MAGSAC++, a fast, reliable and accurate robust estimatorcited by 160
- 2016Deep Image Homography Estimationcited by 135
- 2019SOSNet: Second Order Similarity Regularization for Local Descriptor Learningcited by 131
- 2018MAGSAC: marginalizing sample consensuscited by 122
- 2017Unsupervised Deep Homography: A Fast and Robust Homography Estimation Modelcited by 75
10 more
- 2020Deep Homography Estimation for Dynamic Scenescited by 60
- 2021Unsupervised Deep Image Stitching: Reconstructing Stitched Features to Imagescited by 49
- 2021SOLD2: Self-supervised Occlusion-aware Line Description and Detectioncited by 45
- 2022Iterative Deep Homography Estimationcited by 45
- 2020Deep Hough-Transform Line Priorscited by 44
- 2021Deep Lucas-Kanade Homography for Multimodal Image Alignmentcited by 38
- 2021Motion Basis Learning for Unsupervised Deep Homography Estimation with Subspace Projectioncited by 28
- 2021LocalTrans: A Multiscale Local Transformer Network for Cross-Resolution Homography Estimationcited by 28
- 2021Depth-Aware Multi-Grid Deep Homography Estimation with Contextual Correlationcited by 28
- 2022Unsupervised Homography Estimation with Coplanarity-Aware GANcited by 19
Built on by
Citing papers per year: 2025: 3.
3 papers.