skip to main content
10.1145/3655532.3655554acmotherconferencesArticle/Chapter ViewAbstractPublication PagesicrsaConference Proceedingsconference-collections
research-article

Doc-Former: A transformer-based document shadow denoising network

Published: 28 June 2024 Publication History

Abstract

The existence of shadows makes the visual perception and readability of document images poor, so how to remove the shadows in these document images is an urgent problem to be solved in the industry. Currently, only a few methods are specifically designed for shadow removal of document images. Among them, some algorithms are heuristic algorithms based on experience or direct observation. These algorithms only heuristically denoise the image from the perspective of light or color, and do not take into account the specific characteristics of the shadow of the document. So we propose a transformer-based document shadow denoising algorithm, and the experimental comparison proves that it has achieved state-of-the-art excellence in its performance.

References

[1]
Steve Bako, Soheil Darabi, Eli Shechtman, Jue Wang, Kalyan Sunkavalli, and Pradeep Sen. 2016 . Removing shadows from images of documents. In Proceedings of Asian Confer- ence on Computer Vision (ACCV), pages 173–183.
[2]
Netanel Kligler, Sagi Katz, and Ayellet Tal. 2018. Document en- hancement using visibility detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion (CVPR), pages 2374–2382.
[3]
Seungjun Jung, Muhammad Abul Hasan, and Changick Kim. 2018. Water-filling: An efficient algorithm for digitized doc- ument shadow removal. In Proceedings of Asian Conference on Computer Vision (ACCV), pages 398–414.
[4]
Jifeng Wang, Xiang Li, and Jian Yang. 2018. Stacked conditional generative adversarial networks for jointly learning shadow detection and shadow removal. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 1788–1797.
[5]
Shengfeng He, Bing Peng, Junyu Dong, and Yong Du. 2021. Mask-shadownet: Toward shadow removal via masked adaptive instance normalization. IEEE Signal Process- ing Letters, vol. 28, pp. 957–961.
[6]
Y. -H. Lin, W. -C. Chen and Y. -Y. Chuang. 2020. BEDSR-Net: A Deep Shadow Removal Network From a Single Document Image. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 2020, pp. 12902-12911.
[7]
Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U- Net: Convolutional networks for biomedical image segmen- tation. In Proceedings of International Conference on Med- ical image Computing and Computer-Assisted Intervention (MICCAI), pages 234–241. Springer.
[8]
Mehdi Mirza and Simon Osindero. 2014. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784.
[9]
Li, J., Cheng, B., Chen, Y., Gao, G., & Zeng, T. 2023. EWT: Efficient Wavelet-Transformer for Single Image Denoising. ArXiv, abs/2304.06274.
[10]
Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby .2021. An image is worth 16x16 words: Transformers for image recognition at scale. ICLR.
[11]
ZhendongWang,XiaodongCun,JianminBao,andJianzhuang Liu. 2021. Uformer: A general u-shaped transformer for image restoration. arXiv preprint 2106.03106.
[12]
Zamir S W, Arora A, Khan S, 2022. Restormer: Efficient transformer for high-resolution image restoration[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022: 5728- 5739.

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Other conferences
ICRSA '23: Proceedings of the 2023 6th International Conference on Robot Systems and Applications
September 2023
335 pages
ISBN:9798400708039
DOI:10.1145/3655532
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 28 June 2024

Permissions

Request permissions for this article.

Check for updates

Author Tags

  1. document shadow denoising algorithm
  2. transformer

Qualifiers

  • Research-article
  • Research
  • Refereed limited

Conference

ICRSA 2023

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • 0
    Total Citations
  • 15
    Total Downloads
  • Downloads (Last 12 months)15
  • Downloads (Last 6 weeks)5
Reflects downloads up to 20 Feb 2025

Other Metrics

Citations

View Options

Login options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

HTML Format

View this article in HTML Format.

HTML Format

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media