Development of an Algorithm for Recognition of the Kazakh-Latin Alphabet in the Image
The work is devoted to the study of the alphabet recognition process in the image. The goal is to implement a neural network based on the architecture of convolutional neural networks for recognition of the Kazakh Latin alphabet neural network in practice is considered. A convolutional neural network was formed, which was trained to distinguish handwritten English letters and numbers, since specific symbols and graphic features are a distinctive feature for the Kazakh-Latin alphabet. A neural network model has been constructed for preprocessing a data set. Based on the model, a test data set is synthesized, which is used to evaluate and control the quality of the model during training, and a test data set intended for final evaluation of the model's performance after the training process is completed. The general formulation of an effective algorithm capable of automatically recognizing the symbols of the Kazakh-Latin alphabet in images and presenting them in text format is presented. A new approach based on the use of deep convolutional neural networks is proposed.
Сведения
- Журнал
- Trudy Universiteta
- Опубликовано
- 2023 · Т. 4
- Язык
- Английский
- ISSN
- 1609-1825 · 2710-3382
- Издатель
- Karaganda Technical University named after Abylkas Saginov