Next Article in Journal
Spatiotemporal Characteristics of Radio Frequency Dielectric Barrier Glow Discharge at Atmospheric Pressure
Previous Article in Journal
Combined Use of 3D and HSI for the Classification of Printed Circuit Board Components
Article

HybridTabNet: Towards Better Table Detection in Scanned Document Images

1
Department of Computer Science, Technical University of Kaiserslautern, 67663 Kaiserslautern, Germany
2
Mindgarage, Technical University of Kaiserslautern, 67663 Kaiserslautern, Germany
3
German Research Institute for Artificial Intelligence (DFKI), 67663 Kaiserslautern, Germany
4
Department of Computer Science, Luleå University of Technology, 971 87 Luleå, Sweden
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Academic Editor: Antonio Fernández
Appl. Sci. 2021, 11(18), 8396; https://doi.org/10.3390/app11188396
Received: 16 August 2021 / Revised: 3 September 2021 / Accepted: 3 September 2021 / Published: 11 September 2021
Tables in document images are an important entity since they contain crucial information. Therefore, accurate table detection can significantly improve the information extraction from documents. In this work, we present a novel end-to-end trainable pipeline, HybridTabNet, for table detection in scanned document images. Our two-stage table detector uses the ResNeXt-101 backbone for feature extraction and Hybrid Task Cascade (HTC) to localize the tables in scanned document images. Moreover, we replace conventional convolutions with deformable convolutions in the backbone network. This enables our network to detect tables of arbitrary layouts precisely. We evaluate our approach comprehensively on ICDAR-13, ICDAR-17 POD, ICDAR-19, TableBank, Marmot, and UNLV. Apart from the ICDAR-17 POD dataset, our proposed HybridTabNet outperformed earlier state-of-the-art results without depending on pre- and post-processing steps. Furthermore, to investigate how the proposed method generalizes unseen data, we conduct an exhaustive leave-one-out-evaluation. In comparison to prior state-of-the-art results, our method reduced the relative error by 27.57% on ICDAR-2019-TrackA-Modern, 42.64% on TableBank (Latex), 41.33% on TableBank (Word), 55.73% on TableBank (Latex + Word), 10% on Marmot, and 9.67% on the UNLV dataset. The achieved results reflect the superior performance of the proposed method. View Full-Text
Keywords: table detection; table localization; deep learning; hybrid task cascade; object detection; deformable convolution; deep neural networks; computer vision; scanned document images; document image analysis table detection; table localization; deep learning; hybrid task cascade; object detection; deformable convolution; deep neural networks; computer vision; scanned document images; document image analysis
Show Figures

Figure 1

MDPI and ACS Style

Nazir, D.; Hashmi, K.A.; Pagani, A.; Liwicki, M.; Stricker, D.; Afzal, M.Z. HybridTabNet: Towards Better Table Detection in Scanned Document Images. Appl. Sci. 2021, 11, 8396. https://doi.org/10.3390/app11188396

AMA Style

Nazir D, Hashmi KA, Pagani A, Liwicki M, Stricker D, Afzal MZ. HybridTabNet: Towards Better Table Detection in Scanned Document Images. Applied Sciences. 2021; 11(18):8396. https://doi.org/10.3390/app11188396

Chicago/Turabian Style

Nazir, Danish, Khurram A. Hashmi, Alain Pagani, Marcus Liwicki, Didier Stricker, and Muhammad Z. Afzal. 2021. "HybridTabNet: Towards Better Table Detection in Scanned Document Images" Applied Sciences 11, no. 18: 8396. https://doi.org/10.3390/app11188396

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop