Evaluación de una plataforma completa para Reconocimiento de Textos Manuscritos en Español

• Autores: Eduardo Xamena, Héctor Barbosa, Carlos Ismael Orozco
• Localización: Ciencia y tecnología, ISSN 1850-0870, ISSN-e 2344-9217, Nº. 21, 2021
• Idioma: español
• DOI: 10.18682/cyt.vi21.4327
• Títulos paralelos:
  • End-to-end platform evaluation for Spanish Handwritten Text Recognition
• Enlaces
  • Texto completo (pdf)
• Resumen
  • español

    La tarea del reconocimiento automatizado de textos manuscritos requiere de diversas fases y tecnologías tanto ópticas como del lenguaje. En este artículo se describe un enfoque para la realización de esta tarea de forma completa, mediante el empleo de aprendizaje automatizado a lo largo de todas las fases del proceso. Además de explicar la metodología empleada, se describe el proceso de construcción y evaluación de un modelo de reconocimiento de manuscritos para el lenguaje español. La contribución original de este artículo está dada por el entrenamiento y evaluación de modelos de Offline HTR para manuscritos en español, así como la evaluación de una plataforma para la realización de esta tarea de forma completa. Además, se detallan los trabajos que se están llevando a cabo para lograr mejoras en los modelos obtenidos, y desarrollar nuevos modelos para distintos corpus de lectura compleja.

  • English

    The task of automated recognition of handwritten texts requires various phases and technologies both optical and language related. This article describes an approach for performing this task in a comprehensive manner, using machine learning throughout all phases of the process. In addition to the explanation of the employed methodology, it describes the process of building and evaluating a model of manuscript recognition for the Spanish language. The original contribution of this article is given by the training and evaluation of Offline HTR models for Spanish language manuscripts, as well as the evaluation of a platform to perform this task in a complete way. In addition, it details the work being carried out to achieve improvements in the models obtained, and to develop new models for different complex corpora that are more difficult for the HTR task.

• Referencias bibliográficas
  • Ahlawat, S. & Rishi, R. (2017), Off-line handwritten numeral recognition using hybrid feature set–a comparative analysis, Procedia computer...
  • Bluche, T. (2015), Deep neural networks for large vocabulary handwritten text recognition, PhD thesis, Paris 11.
  • Castro, D.; Bezerra, B. L. D. & Valença, M. (2018), Boosting the deep multidimensional long-short-term memory network for handwritten...
  • De Sousa Neto, A. F.; Bezerra, B. L. D.; Toselli, A. H. & Lima, E. B. (2020), HTR-Flor++: A Handwritten Text Recognition System...
  • Granell, E.; Romero, V. & Martinez-Hinarejos, C.-D. (2020), Study of the influence of lexicon and language restrictions on computer assisted...
  • Graves, A.; Fernández, S.; Gomez, F. & Schmidhuber, J. (2006), Connectionist temporal classification: labelling unsegmented sequence data...
  • Graves, A. & Schmidhuber, J. (2009), Offline handwriting recognition with multidimensional recurrent neural networks, in Advances in neural...
  • Grüning, T., Leifert, G., Strauß, T., Michael, J., & Labahn, R. (2019). A two-stage method for text line detection in historical documents....
  • Jeong, J.; Park, H. & Kwak, N. (2017), Enhancement of SSD by concatenating feature maps for object detection, CoRR abs/1705.09587.
  • Kang, L.; Riba, P.; Rusiсol, M.; Fornés, A. & Villegas, M. (2020), Pay Attention to What You Read: Non-recurrent Handwritten Text-Line...
  • Kar, R.; Saha, S.; Bera, S. K.; Kavallieratou, E.; Bhateja, V. & Sarkar, R. (2019), Novel approaches towards slope and slant correction...
  • Liebl, B., & Burghardt, M. (2020). An Evaluation of DNN Architectures for Page Segmentation of Historical Newspapers. arXiv preprint arXiv:2004.07317.
  • Michael, J.; Labahn, R.; Grüning, T. & Zöllner, J. (2019), Evaluating sequenceto-sequence models for handwritten text recognition, in...
  • More, P. K. & Dighe, D. D. (2016), A review on document image binarization technique for degraded document images, Int. Res. J. Eng. Technol,...
  • Neto, A. F. S.; Bezerra, B. L. D. & Toselli, A. A. H. (2020), Towards the Natural Language Processing as Spelling Correction for Offline...
  • Niblack, W. (1986), An Introduction to Digital Image Processing (Englewood CliVs, NJ, Prentice-Hall.
  • Oliveira, S. A.; Seguin, B. & Kaplan, F. (2018), dhSegment: A generic deeplearning approach for document segmentation, in 2018 16th International...
  • Otsu, N. (1979), A threshold selection method from gray-level histograms, IEEE transactions on systems, man, and cybernetics 9(1), 62--66.
  • Reul, C.; Christ, D.; Hartelt, A.; Balbach, N.; Wehner, M.; Springmann, U.; Wick, C.; Grundig, C.; Büttner, A. & Puppe, F. (2019), OCR4all—An...
  • Romero, V.; Fornés, A.; Serrano, N.; Sánchez, J. A.; Toselli, A. H.; Frinken, V.; Vidal, E. & Lladуs, J. (2013), The ESPOSALLES database:...
  • Sánchez, J. A.; Mühlberger, G.; Gatos, B.; Schofield, P.; Depuydt, K.; Davis, R. M.; Vidal, E. & de Does, J. (2013), tranScriptorium:...
  • Sánchez, J. A.; Romero, V.; Toselli, A. H.; Villegas, M. & Vidal, E. (2019), A set of benchmarks for handwritten text recognition on historical...
  • Sarathy, S. & Manikandan, J. (2018), Design and evaluation of a real-time character recognition system, in 2018 International Conference...