Abstract:

Sign language is the movements they make simultaneously with face expressions to communicate the thoughts of the speaker with the orientation of the movements of the hands, fingers, arms or body movements. Signal languages have become the favourite position of all researchers in recent years. Movements can be recognized with the help of sensors. However, the transfer of movement data to computer systems is of great importance. The results of the summer study have found that the work in this direction is not sufficient. Furthermore, the studies were found that more studies were conducted in the direction of American Sign Language, English Sign Language and Arabic Sign Language, and that the studies were conducted in the direction of Turkish Sign Language were not sufficient. In this study, an intelligent system was developed that could facilitate the communication of hearing and speaking disabled individuals with other individuals. It is believed that the work done in this context will benefit from removing this deficiency in the field. In the smart system developed within the framework of the study, 33 basic signals, called the sound science in the Turkish Sign Language, were founded as the basis of the signals. The Microsoft Kinect v2 sensor has been used to recognize these signs by the system. In the system’s infrastructure, the C# programming language and classification algorithms are hidden Markov Model and MongoDB as a database. The case study found that 82% of 33 core signs were correctly identified by the developed system. It is believed that the signal recognition system developed taking into account the accuracy ratio achieved will help both hearing and speech disabled individuals and other individuals and will solve the problem of communicating between them.

Keywords: