Military Hand Signal Recognition System with Digital Image Processing

Abstract

Today's military hand signals are a form of communication used to communicate with each other. It is a signal that can be understood through the eyes. It is any communication method that can be used to forward signal. And it is a fast sending of messages over a short distance, which the visual signal of this nature has some limitations, for example, communication range and reliability are significantly reduced during visibility problems. and when the terrain is limited. At present, the personnel in the army, although there is always training and review of hand signals. And it is also a training by the strength of a person to train a person. But because of the current Covid-19 outbreak and the country has entered the Thailand 4.0 era, which focuses on developing technology with creativity and developing various innovations for the benefit of the nation. Based on the above concepts, it is therefore an important factor to develop personnel in the area related to technology by training personnel with computer programs. It starts with learning to recognize hand gestures using a computer with digital image processing. Perform hand gesture detection and recognition. From the algorithm, the number of 3 methods is Neural Network, K-Nearest Neighbour (K-NN) and Template Matching. The results showed that the neural network method yielded the most accurate results at 82.326%.