Abstract:Homography estimation refers to the problem of computing a 3 × 3 matrix which transfers image points between two images of a planar scene or two images captured from the same location. While existing algorithms exploiting hand-crafted sparse image features are well-established and efficient, recent methods based on convolutional neural networks (CNNs) achieve promising results especially for low-texture scenes. This work proposes to solve homography estimation using a hybrid framework HomoNetComb which incorporates deep learning method and energy minimization. In particular, a customized light-weight CNN named HomoNetSim is designed to calculate an initial estimation of homography, where the network is trained in an end-to-end fashion using large amount of image pairs generated from a publicly available dataset. Due to the tiny size of the employed network, the computation time of both training and inference for HomoNetSim can be reduced significantly compared with existing CNN-based homography estimation method. The initial estimate is then refined via gradient-decent algorithm by minimizing the masked pixel-level photometric discrepancy between the warped image and the destination image in a parallel fashion. Extensive experiments on the large scale synthetic dataset demonstrate that the proposed HomoNetComb improves robustness of homography estimation significantly compared with traditional methods based on sparse image features, and meanwhile HomoNetComb achieves a mean average corner error (MACE) of 0.58 pixels which outperforms previous state-of-the-art CNN-based method. Moreover, the usefulness and applicability of the proposed method is demonstrated by applying it to solve a real-world image stitching problem.

Joint Edge Detection Learning for Recurrent Homography Estimation

Recurrent Homography Estimation Using Homography-Guided Image Warping and Focus Transformer

Edge-aware Correlation Learning for Unsupervised Progressive Homography Estimation

MCNet: Rethinking the Core Ingredients for Accurate and Efficient Homography Estimation

Combining Convolutional Neural Network and Photometric Refinement for Accurate Homography Estimation.

Content-Aware Unsupervised Deep Homography Estimation and its Extensions

RTHEN: Unsupervised deep homography estimation based on dynamic attention for repetitive texture image stitching

Deep Homography Estimation Via Dense Scene Matching

Homography Estimation Network Based on Dense Correspondence

Learning edge-preserved image stitching from multi-scale deep homography

Learning Edge-Preserved Image Stitching from Large-Baseline Deep Homography

Homography Decomposition Networks for Planar Object Tracking

A highly accurate dense approach for homography estimation using modified differential evolution.

Image stitching via deep homography estimation

AbHE: All Attention-Based Homography Estimation

Unsupervised Global and Local Homography Estimation with Motion Basis Learning

SRHEN: Stepwise-Refining Homography Estimation Network via Parsing Geometric Correspondences in Deep Latent Space

Unsupervised deep homography with multi-scale global attention.

Unsupervised Deep Homography: A Fast and Robust Homography Estimation Model

HEASK: Robust Homography Estimation Based on Appearance Similarity and Keypoint Correspondences.

CodingHomo: Bootstrapping Deep Homography with Video Coding