Research of Stages, Types of Modeling and Methods of Gesture Recognition

Abstract

It is impossible to leave aside the problem of communication with the outside world and the social integration of people with disabilities in modern society, where there is a rapid development of communication methods between people. Gesture is the only method of communication for people with disabilities (auditory and language). The research provides information on static and dynamic gestures that allow people with communication problems to exchange information. It is noted that the problem of automatic gesture recognition is solved by various mathematical methods, algorithms and computer systems. The advantages and disadvantages of 2D and 3D models for sequential recognition of hand gestures are shown. It is noted that the creation of a national electronic database that recognizes the national dactyl alphabet and gestures in the Republic of Azerbaijan is an urgent problem. The well-known methods of this category are shown. Artificial neural network method, Hidden Markov method, Random Forest method and method of designating a hand gesture using gloves painted with markers of different colors. The creation of a national electronic database that recognizes the national dactyl alphabet and gestures is proposed, which will partially eliminate the existing barrier in education and problems of communication between people.