Research and development of image processing algorithms for effective recognition of various gestures in real time

Abstract

The study proposed a method of sign language communication using machine learning. Various sign language standards are considered. The study uses neural networks for gesture recognition, namely Convolutional Neural Networks (CNN). The work will also use OpenCV technology to capture gestures from video. The considered dataset for the neural network is ASL Alphabet. An overview of neural networks is provided, as well as a detailed description of how to apply and build a convolutional neural network. It is indicated by what means and where the development of the software product took place. The libraries that were used to perform the given task are described. The architecture of the convolutional neural network, which was used in the implementation of the software product, is indicated. After training, the neural network was tested and showed an accuracy of 90.16%. The software product is described and a user manual is created.