THE INVESTIGATION OF THE HANDWRITING STROKES MODEL
Abstract
In this paper we propose and study a model of handwriting strokes. The model
consider handwritten text as a collection of strokes of several known classes. Model
allows us to abstract from the initial conditions of the writing, such as angle, velocity,
density, etc. Thanks to the procedure of tracing the model is irrelevant to noise of the
original image and the way it was received. We conducted an experiment, confirming the
applicability of the model for use in pattern recognition problems
consider handwritten text as a collection of strokes of several known classes. Model
allows us to abstract from the initial conditions of the writing, such as angle, velocity,
density, etc. Thanks to the procedure of tracing the model is irrelevant to noise of the
original image and the way it was received. We conducted an experiment, confirming the
applicability of the model for use in pattern recognition problems
Keywords
neural network,
mathematical model,
python,
recognition,
vector,
count,
language model,
recognition,
stroke
About the Authors
Е. Долгова
Пермский государственный технический университет
Russian Federation
Russian Federation
Д. Курушин
Пермский государственный технический университет
Russian Federation
Russian Federation
References
1. Долгова, Е.В., Курушин Д.С. Компьютерные нейросетевые технологии, Пермь, ПГТУ, 2008.
2. Мисюрёв, А.В. Использование искусственных нейронных сетей для распознавания рукопечатных символов, [Электронный документ] (http://ocrai.narod.ru/hp.html). Проверено 2010.12.20.
3. Шаров, С.А. Статистика слов в русском языке. [Электронный документ] (http://www.lingvisto.org/artikoloj/ru_stat.html). Проверено 12.02.2011.
4. Jaehwa, Park, Venu Govindaraju, and Sargur N. Srihari. Efficient word segmentation driven by unconstrained handwritten phrase recognition. In Proceedings of International Conference on Document Analysis and Recognition, pages 605-608, 1999.
5. Selinger, P. Potrace: a polygon-based tracing algorithm языке. [Электронный документ] (http://potrace.sourceforge.net/potrace.pdf). Проверено 12.02.2011.
6. Райер, И.А. Сегментация штрихов и их соединений при распознавании рукописного текста / Материалы Международной конференции по компьютерной графике и машинному зрению. Ч.III: С.151-155, Москва, 1999.
7. Абраменко, А. Компьютер читает [Электронный ресурс]. Электрон. текстовые дан. - 2000. Режим доступа: http://www.ocrai.narod.ru.
8. Кузнецов, А.В. Распознавание рукописного текста [Электронный ресурс]. Электрон. текстовые дан. - 2001. Режим доступа: http://www.masters.donntu.edu.ua/2001/fvti/kuznetsov/diss/lib/aiconcep/rec_text.htm
9. Suen, H.M., Wang, J.F. «Segmentation of Uniform Colored Text from Color Graphics Background», VISP-144, 1997, P. 317-322.
Review
Views:
60
Email this article 