Unsupervised Deep Image Stitching: Reconstructing Stitched Features to Images

@article{Nie2021UnsupervisedDI,
  title={Unsupervised Deep Image Stitching: Reconstructing Stitched Features to Images},
  author={Lang Nie and Chunyu Lin and Kang Liao and Shuaicheng Liu and Yao Zhao},
  journal={IEEE Transactions on Image Processing},
  year={2021},
  volume={30},
  pages={6184-6197}
}
Traditional feature-based image stitching technologies rely heavily on feature detection quality, often failing to stitch images with few features or low resolution. The learning-based image stitching solutions are rarely studied due to the lack of labeled data, making the supervised methods unreliable. To address the above limitations, we propose an unsupervised deep image stitching framework consisting of two stages: unsupervised coarse image alignment and unsupervised image reconstruction… Expand
Depth-Aware Multi-Grid Deep Homography Estimation with Contextual Correlation
TLDR
This paper designs a contextual correlation layer, which can capture the long-range correlation on feature maps and flexibly be bridged in a learning framework, and proposes to predict multi-grid homography from global to local. Expand

References

SHOWING 1-10 OF 53 REFERENCES
Learning Edge-Preserved Image Stitching from Large-Baseline Deep Homography
TLDR
An image stitching learning framework, which consists of a large-baseline deep homography module and an edge-preserved deformation module, which is superior to the existing learning method and shows competitive performance with state-of-the-art traditional methods. Expand
A view-free image stitching network based on global homography
TLDR
A cascaded view-free image stitching network based on a global homography, which can achieve almost 100% elimination of artifacts in overlapping areas at the cost of acceptable slight distortions in non-overlapping areas, compared with traditional methods. Expand
Deep Feature Extraction for Panoramic Image Stitching
TLDR
Different from most existing image matching methods, the proposed method extracts image features using deep learning approach directly estimates locations of features between pairwise constraint of images by maximizing an image- patch similarity metric between images. Expand
Warping Residual Based Image Stitching for Large Parallax
  • Kyu-Yul Lee, Jae-Young Sim
  • Computer Science
  • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2020
TLDR
This paper proposes an image stitching algorithm robust to large parallax based on the novel concept of warping residuals and partitions input images into superpixels and warp each superpixel adaptively according to an optimal homography which is computed by minimizing the error of feature matches weighted by the warping remnants. Expand
Deep Image Homography Estimation
TLDR
Two convolutional neural network architectures are presented for HomographyNet: a regression network which directly estimates the real-valued homography parameters, and a classification network which produces a distribution over quantized homographies. Expand
Parallax-Tolerant Image Stitching
  • Fangfang Zhang, F. Liu
  • Computer Science
  • 2014 IEEE Conference on Computer Vision and Pattern Recognition
  • 2014
TLDR
This paper presents a local stitching method to handle parallax based on the observation that input images do not need to be perfectly aligned over the whole overlapping region for stitching, and develops a seam finding method that estimates a plausible seam from only roughly aligned images by considering both geometric alignment and image content. Expand
Real-Time Image Stitching with Convolutional Neural Networks
TLDR
A novel approach is proposed to capture the images to be stitched and the corresponding ground truth, and make a dataset generated by the proposed method, which will make progresses for real time image stitching. Expand
Multi-scenes Image Stitching Based on Autonomous Driving
  • Lang Wang, Wen Yu, Bao Li
  • Computer Science
  • 2020 IEEE 4th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC)
  • 2020
TLDR
A multi-scenes image stitching method for autonomous driving, which uses autoencoder networks to extract feature points and includes a dimensionality-reduced feature extraction path and a precisely located symmetrical decoding path. Expand
Parallax-Tolerant Image Stitching Based on Robust Elastic Warping
TLDR
A parallax-tolerant image stitching method based on robust elastic warping, which could achieve accurate alignment and efficient processing simultaneously and is highly compatible with different transformation types. Expand
Adaptive as-natural-as-possible image stitching
TLDR
A novel stitching method, that uses a smooth stitching field over the entire target image, while accounting for all the local transformation variations, that is more robust to parameter selection, and hence more automated compared with state-of-the-art methods. Expand
...
1
2
3
4
5
...