Recognition of Handwritten Texts on Images Using Deep Machine Learning-Reference-Cited by-同舟云学术

Recognition of Handwritten Texts on Images Using Deep Machine Learning

Published:2024 Issue:1 (305) Volume: Page:50-56
ISSN:2706-8145
Container-title:Control Systems and Computers
language:
Short-container-title:Control syst. comput.

Author:

,Snitko Marianna D.,Khitsko Iana V.^ORCID, ,Rybachok Nataliia A.^ORCID,

Abstract

The article is devoted to the aspects of using deep machine learning to recognize handwritten text containing letters of the Latin alphabet and numbers. Software has been developed that recognizes handwritten text. A convolutional neural network consisting of 13 layers was trained for 50 epochs on images of 814255 characters taken from the EMNIST dataset. The prediction accuracy was 0.9468, the response rate was 0,9673, the F1-index reached 0,9429, and the average processing time of one image was 1,15 seconds.

Publisher

National Academy of Sciences of Ukraine (Co. LTD Ukrinformnauka) (Publications)

Reference11 articles.

1. 1. LeCun, Y., Cortes, C., Burges, C.J.C. "The MNIST Database of Handwritten Digits" [online]. Available at: [Accessed 06 Dec. 2023].

2. 2. Digital Image Processing / Rafael C. González; Richard Eugene Woods (ed.). Prentice Hall, 2007, pp. 85, ISBN 978-0-13-168728-8.

3. 3. Chauhan, S., Sharma, E., Doegar, A. (2016). "Binarization Techniques for Degraded Document Images. A Review". 5th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO).

4. 4. Krizhevsky, A., Sutskever, I., Hinton, G.E. (2012). ImageNet Classification with Deep Convolutional Neural Networks. Advances in Neural Information Processing Systems 25 (NIPS 2012), pp. 1097-1105.

5. 5. Mouton, C., Myburgh, J.C., & Davel, M.H. (2020). "Stride and Translation Invariance in CNNS". In Southern African Conference for Artificial Intelligence Research, Vol. 1342, Cham: Springer International Publishing, pp. 267-281. arXiv:2103.10097. doi:10.1007/978-3-030-66151- 9_17.