American sign language recognition: algorithm and experimentation system

Abstract

The objective of this paper is an attempt to take part in the fast development of computer science, especially machine learning, in the field of finding the tools for supporting the disabled people. Speech-and–hearing impaired people are the part of our society and it would be a great convenience both for them and for speaking people, who would have an opportunity for a better communication using computer technology. In this paper, the results of the research concerning recognition of sign language have been provided. The research includes experimenting with the images transformations and the usage of different learning and feature detecting algorithms to obtain the best quality of signs recognition. In addition, the recommendations based on results of experiments can be applied in practical issues, for example, in determining hands rotations used in sign language, which can improve the accuracy of recognition.