Распознавание табличной информации с использованием свёрточных нейронных сетей

Бесплатный доступ

Показана актуальность выявления табличной информации и распознавания её содержимого для обработки отсканированных документов. Описано формирование набора данных для обучения, валидации и тестирования нейронной сети глубокого обучения (DNN) YOLOv5s для обнаружения простых таблиц. Отмечена эффективность использования этой DNN при работе с отсканированными документами. С использованием Keras Functional API сформирована свёрточная нейронная сеть (CNN) для распознавания основных элементов табличной информации - цифр, основных знаков препинания и букв кириллицы. Приведены результаты исследования работы этой CNN. Описана реализация выявления и распознавания табличной информации на отсканированных документах в разработанной ИС актуализации информации в базах данных системы ЕГРН Росреестра.

Еще

Свёрточные нейронные сети, нейронные сети глубокого обучения, cnn, dnn, yolov5s, keras, python

Короткий адрес: https://sciup.org/143180112

IDR: 143180112   |   DOI: 10.25209/2079-3316-2023-14-1-3-30

Список литературы Распознавание табличной информации с использованием свёрточных нейронных сетей

  • I. V. Vinokurov. "Using a convolutional neural network to recognize text elements in poor quality scanned images", Programmny'e sistemy': teoriya i prilozheniya, 13:3 (2022), pp. 29-43 (in Russian). |gl url; 'tis 22 23 26 28
  • G. Harit, A. Bansal. "Table detection in document images using header and trailer patterns", Proceedings of the Eighth Indian Conference on Computer Vision, Graphics and Image Processing, ICVGIP'12 (December 16-19, 2012, Mumbai, India), ACM, New York, 2012, ISBN 978-1-4503-1660-6, 8 pp. i 19
  • B. Gatos, D. Danatsas, I. Pratikakis, S. Perantonis. "Automatictable detection in document images", ICAPR 2005: Pattern Recognition and Data Mining, Lecture Notes in Computer Science, vol. 3686, Springer, Berlin-Heidelberg, 2005, ISBN 978-3-540-28757-5, pp. 609-618. i 19
  • T. Kasar, P. Barlas, S. Adam, C. Chatelain, T. Paquet. "Learning to detect tables in scanned document images using line information", 2013 12th International Conference on Document Analysis and Recognition (25-28 August 2013, Washington, DC, USA), 2013, pp. 1185-1189. i 119
  • M. A. Jahan, R. G. Ragel. "Locating tables in scanned documents for reconstructing and republishing", 7th International Conference on Information and Automation for Sustainability (22-24 December 2014, Colombo, Sri Lanka), 2014, pp. 1-6. 19
  • T. G. Kieninger. "Table structure recognition based on robust block segmentation", Document Recognition V, Photonics West'98 Electronic Imaging (1998, San Jose, CA, United States), Proc. SPIE, vol. 3305, 1998, pp. 22-32. d fig
  • Li Minghao, Cui Lei, Huang Shaohan, Wei Furu, Zhou Ming, Li Zhoujun. TableBank: A benchmark dataset for table detection and recognition, 2020, 9 pp.
  • J. Fang, X. Tao, Z. Tang, R. Qiu, Y. Liu. "Dataset, ground-truth and performance metrics for table detection evaluation", 2012 10th IAPR International Workshop on Document Analysis Systems (27-29 March 2012, Gold Coast, QLD, Australia), 2012, pp. 445—449. 20
  • M. Gobel, T. Hassan, E. Oro, G. Orsi. "Icdar 2013 table competition", 2013 12th International Conference on Document Analysis and Recognition (15 October 2013, Washington, DC, USA), 2013, pp. 1449-1453. 20
  • A. Shahab, F. Shafait, T. Kieninger, A. Dengel. "An open approach towards the benchmarking of table structure recognition systems", Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS'10 (June 9-11, 2010, Boston, Massachusetts, USA), 2010, ISBN 978-1-60558-773-8, pp. 113-120. 20
  • L. Gao, Y. Huang, H. Dejean, J.-L. Meunier, Q. Yan, Y. Fang, F. Kleber, E. Lang. "ICDAR 2019 competition on table detection and recognition (cTDaR)", 2019 International Conference on Document Analysis and Recognition (ICDAR) (20-25 September 2019, Sydney, NSW, Australia'), 2019, pp. 1510-1515. 20
  • S. Ren, K. He, R. Girshick, J. Sun. "Faster R-CNN: towards real-time object detection with region proposal networks", IEEE transactions on pattern analysis and machine intelligence, 39:6 (2016), pp. 1137-1149. 20
  • J. Redmon, S. Divvala, R. Girshick, A. Farhadi. "You only look once: Unified, real-time object detection", Proceedings of the IEEE conference on computervision and pattern recognition, 2016, pp. 779-788. 20
  • A. Gilani, S. R. Qasim, I. Malik, F. Shafait. "Table detection using deep learning", 201714th IAPR International Conference on Document Analysis and Recognition (ICDAR). V. 1 (09-15 November 2017, Kyoto, Japan), 2017, pp. 771-776. d -f2o
  • A. Banerjee. YOLOv5 vs YOLOv6 vs YOLOv7, Learn With A Robot, https://www.learnwitharobot.eom/p/yolov5-vs-yolov6-vs-yolov7, 2022-2023.|20
  • P. Lebiedzinski. A single number metric for evaluating object detection models, Towards Data Science, https://towardsdatascience.com/a-single-number-metric-for-evaluating-object-detection-models-c97f4a98616d,2021.^20
  • Surya Gutta. Object Detection Algorithm — YOLO v5 Architecture, https://medium.com/analytics-vidhya/object-detection-algorithm-yolo-v5-architecture-89e0a35472ef, Analytics Vidhya, 2021. t"20
  • Zixin Ning, Xinjiao Wu, Jing Yang, Yanqin Yang. "MT-YOLOv5: Mobile terminal table detection model based on YOLOv5", The Fourth International Conference on Physics, Mathematics and Statistics (ICPMS) 2021 (19-21 May 2021, Kunming, China), Journal of Physics: Conference Series, 1978 (2021), 012010. d -f2o
  • Yilun Huang, Qinqin Yan, Yibo Li, Yifan Chen, Zhi Tang. "A YOLO-based table detection method", 2019 International Conference on Document Analysis and Recognition (ICDAR) (20-25 September 2019, Sydney, NSW, Australia), 2019.
Еще
Статья научная