Abstract:Irregular boundaries in image stitching naturally occur due to freely moving cameras. To deal with this problem, existing methods focus on optimizing mesh warping to make boundaries regular using the traditional explicit solution. However, previous methods always depend on hand-crafted features (e.g., keypoints and line segments). Thus, failures often happen in overlapping regions without distinctive features. In this paper, we address this problem by proposing RecStitchNet, a reasonable and effective network for image stitching with rectangular boundaries. Considering that both stitching and imposing rectangularity are non-trivial tasks in the learning-based framework, we propose a three-step progressive learning based strategy, which not only simplifies this task, but gradually achieves a good balance between stitching and imposing rectangularity. In the first step, we perform initial stitching by a pre-trained state-of-the-art image stitching model, to produce initially warped stitching results without considering the boundary constraint. Then, we use a regression network with a comprehensive objective regarding mesh, perception, and shape to further encourage the stitched meshes to have rectangular boundaries with high content fidelity. Finally, we propose an unsupervised instance-wise optimization strategy to refine the stitched meshes iteratively, which can effectively improve the stitching results in terms of feature alignment, as well as boundary and structure preservation. Due to the lack of stitching datasets and the difficulty of label generation, we propose to generate a stitching dataset with rectangular stitched images as pseudo-ground-truth labels, and the performance upper bound induced from the it can be broken by our unsupervised refinement. Qualitative and quantitative results and evaluations demonstrate the advantages of our method over the state-of-the-art.

Deep Image Stitching with Pixel Similarity Correlation

Remote Sensing Image Stitch Using Modified Structure Deformation

A Robust Image Stitching Based on Solution Set Smoothing

RecStitchNet: Learning to Stitch Images with Rectangular Boundaries

Image stitching via deep homography estimation

Pixel-wise Deep Image Stitching

A view-free image stitching network based on global homography

Learning Pixel-wise Alignment for Unsupervised Image Stitching

Unsupervised Deep Image Stitching: Reconstructing Stitched Features to Images

Image Stitching Based on Semantic Planar Region Consensus

Parallax-Tolerant Unsupervised Deep Image Stitching

Research on Image Stitching Algorithm Based on Point-Line Consistency and Local Edge Feature Constraints

Image Stitching of Scenes with Large Misregistration

UAV remote sensing image stitching via improved VGG16 Siamese feature extraction network

Image Stitching Based on Planar Region Consensus

Pixel-Wise Warping for Deep Image Stitching

Learning Thin-Plate Spline Motion and Seamless Composition for Parallax-Tolerant Unsupervised Deep Image Stitching

Implicit Neural Image Stitching

Content-Seam-Preserving Multi-Alignment Network for Visual-Sensor-Based Image Stitching

SC-AOF: A Sliding Camera and Asymmetric Optical-Flow-Based Blending Method for Image Stitching

Image stitching by line-guided local warping with global similarity constraint