Hand Gesture Recognition System for Arabic letters and numbers Using Deep Learning

Abstract

Recognizing hand gestures is a key to overcome many of the difficulties people with a physical disability have in communicating with the general public. Therefore, it was necessary to develop a technology that solves this problem and enables people with disabilities to communicate with people without problems. This study presents a hand gesture recognition system to recognize Arabic numbers and letters using artificial intelligence that uses deep learning technology. The convolution neural network (CNN) was used as a deep learning model to train data sets on hand gestures, where the gesture images were displayed on the network entrance and changed Scale the images to the same size to extract the image features and then categorize them into text. The results showed that the CNN model achieved an accuracy of 99% in the testing phase.