A comprehensive list of awesome document image rectification methods based on deep learning.
- The results are the rectified images on the DocUNet Benchmark Dataset.
Year | Conf. | Title | Type | Link |
---|---|---|---|---|
2018 | CVPR | DocUNet: Document Image Unwarping via A Stacked U-Net | Real | Dataset |
2019 | TOG | Document Rectification and Illumination Correction using a Patch-based CNN | Synthetic | Dataset |
2019 | ICCV | DewarpNet: Single-Image Document Unwarping With Stacked 3D and 2D Regression Networks | Synthetic | Dataset |
2020 | DAS | Dewarping Document Image by Displacement Flow Estimation with Fully Convolutional Network | Synthetic | Dataset |
2021 | ICDAR | Document Dewarping with Control Points | Synthetic | Dataset |
2022 | CVPR | Fourier Document Restoration for Robust Document Dewarping and Recognition | Real | Dataset |
2022 | SIGGRAPH | Learning From Documents in the Wild to Improve Document Unwarping | Real | Dataset |
2022 | ECCV | Geometric Representation Learning for Document Image Rectification | Real | Dataset |
- This repository is scheduled to be updated regularly in accordance with schedules of major AI conferences.