Segmented And Segmentation Free Approach for Handwritten Text Recognition
Keywords:
HTR (HandWritten Text Recognition), CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), ANN (Artificial Neural Network)Abstract
Text has a long history that dates back thousands of years. In a wide range of vision-based application scenarios, the rich and accurate semantic information carried by text is crucial. As a result, computer vision and pattern recognition researchers have been working on text detection in natural settings.With the growth and development of deep learning in recent years, many techniques have demonstrated promise in terms of originality, viability, and efficiency.In this paper Authors discuss the approach on handwritten text recognition (HTR) with segmentation and without segmentation.For those working in the field of computer vision's text-based picture segmentation, this paper serves as a reference.There are various methods available for segmentation like Histogram , Projection methods. Segmentation is done on 3 levels as lines, word and character. There are also some methods available which do not need segmentation for HTR as well word spotting. These segmentation free methods use different neural network algorithms like CNN, RNN, ANN etc.
Downloads
References
Dave, N. (2015). Segmentation methods for handwritten character recognition. International journal of signal processing, image processing and pattern recognition, 8(4), 155-164.
Fischer, A., Keller, A., Frinken, V., & Bunke, H. (2012). Lexicon-free handwritten word spotting using character HMMs. Pattern recognition letters, 33(7), 934-942.
Patel, C., & Desai, A. (2013). Extraction of characters and modifiers from handwritten Gujarati words. International Journal of Computer Applications, 73(3).
Lemaitre, A., Camillerapp, J., & Coüasnon, B. (2011, January). A perceptive method for handwritten text segmentation. In Document recognition and retrieval XVIII (Vol. 7874, pp. 100-108). SPIE.
Tripathy, N., & Pal, U. (2006). Handwriting segmentation of unconstrained Oriya text. Sadhana, 31, 755-769.
Garg, N. K., Kaur, L., & Jindal, M. K. (2010, April). A new method for line segmentation of handwritten Hindi text. In 2010 seventh international conference on information technology: new generations (pp. 392-397). IEEE.
Lemaitre, A., Camillerapp, J., & Coüasnon, B. (2011, January). A perceptive method for handwritten text segmentation. In Document recognition and retrieval XVIII (Vol. 7874, pp. 100-108). SPIE.
Mistry, N., Vashi, S., Patel, V., Shah, K., Rixawapla, D., Rakholiya, F., & Savant, R. INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY A REVIEW ON SEGMENTATION TECHNIQUES OF LINES, WORDS AND CHARACTERS ON GUJARATI HANDWRITTEN DOCUMENT USING OCR.
Zagoris, K., Amanatiadis, A., & Pratikakis, I. (2021). Word spotting as a service: an unsupervised and segmentation-free framework for handwritten documents. Journal of Imaging, 7(12), 278.
Khaissidi, G., Elfakir, Y., Mrabti, M., El Yacoubi, M., & Chenouni, D. (2016). Segmentation-free word spotting for handwritten Arabic documents.
Chaudhari, S., & Gulati, R. (2014). Segmentation problems in handwritten Gujarati text. International Journal of Engineering Research & Technology (IJERT).
Krishnan, P., & Jawahar, C. V. (2016). Matching handwritten document images. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14 (pp. 766-782). Springer International Publishing.
