Article

Received 18.08.2023

Revised 01.12.2023

Accepted 27.12.2023

Retrieved from Vol. 27, No. 3, 2023

Pages 29 -36

Bezverkhiy, O., & Luz, V. (2023). Designing a sign language recognition information system using OpenCV, TensorFlow and Keras implemented in PYTHON. The National Transport University Bulletin, 27(3), 29-36. https://doi.org/10.33744/2308-6645-2023-3-57-027-034

Designing a sign language recognition information system using OpenCV, TensorFlow and Keras implemented in PYTHON

Oleksandr Bezverkhiy Vladyslav Luz

Sign language is one of the main means of information transfer and communication, along with text and speech. Sign languages consist of individual signs that are combined into letters, words, and phrases using a sequential transition from one sign to another. The article examines and develops the components of Ukrainian sign language translation information technology. The object of the study is the process of dactyl recognition and modeling of the Ukrainian dactyl language. The subject of research is the methods of dactyl image analysis and the process of dactyl recognition and modeling. The purpose of this work is to design an information system for recognizing Ukrainian sign language using Python. This paper examines the integration of OpenCV, TensorFlow, and Keras implemented in Python to develop a robust and accurate sign language recognition system. A detailed and step-by-step explanation of the software code for creating an information system for sign language recognition using OpenCV, TensorFlow and Keras in Python is given The software tools for studying and translating Ukrainian sign language created during the research can be used when designing an information system for translation from one language to another, which will facilitate communication between people with hearing impairments and with those who do not know sign language

sign language; neural networks; information system; python; dactyl alphabet
  1. Kulbida S.V. Ukrainian dactylology: Scientific method manual. – K.: Pedagogical thought, 2007. – 256 p.
  2. Analysis of modern dactyl sign language recognition systems for sign language translation systems / Lykhosherstov D.O., Lebedev D.Yu. // Scientific notes of TNU named after V.I. Vernadskyi. Series: Technical sciences. – Volume 32 (71), No. 6, 2021. – P. 44–48.
  3. Information technology of dactyl identification of Ukrainian sign language / Yu. V. Krak, O. V. Barmak, V. S. Kasyanyuk, D. V. Shkilnyuk // Control systems and machines. – 2015. – No. 6. – P. 23–28.
  4. Information technology for modeling Ukrainian sign language / Yu. G. Kryvonis, Yu. V. Krak, O. V. Barmak et al. // Artificial Intelligence. – 2009. – No. 3. – P. 186–197.
  5. Kondratyuk S.S., Krak Yu.V. Platform-independent software for the development of gestural communication systems: modeling of dactyl speech // Artificial Intelligence. – 2016. – Vol. 73, Art. 3. – P. 36–47.
  6. Sign language as a full-fledged means of communication. Access mode: https://www.naurok.com.ua/zhestova-mova-yak-povnocinniy-zasib-komunikaci-332294.html
  7. OpenCV is a library for recognizing objects in images and cameras. Access mode: https://blog.desdelinux.net/uk/opencv-library-for-object-recognition-in-images-and-cameras/
  8. TensorFlow Core. Documentation. Access mode: https://www.tensorflow.org/api_docs/python/tf
  9. TensorFlow Core. Keras. Access mode: https://www.tensorflow.org/guide/keras/sequential_model
  10. OpenCV-Python Tutorial. Access mode: https://docs.opencv.org/3.4/d6/d00/tutorial_py_root.html
  11. TensorFlow for Beginners With Examples and Python Implementation. Access mode: https://www.analyticsvidhya.com/blog/2021/11/tensorflow-for-beginners-with-examples-and-pythonimplementation/

https://doi.org/10.33744/2308-6645-2023-3-57-027-034