Recognition of sign language using inertial sensors
Recognition of sign language. Aid to communication.
Automatic translation system of sign language signs, providing the possibility of recognising the different signs and creating sentences and words for their translation or reproduction.
The objective was the recognition of sign language using sensorised gloves and inertial sensors.
Through the analysis of the information received from both sensor systems, automatic recognition was carried out in real time of the signs made by the user.
Using various sign processing techniques and artificial intelligence it is possible to segment the temporal signs for subsequent recognition of patterns.
The glove used in this case is made up of 14 flexible sensors that enable characterising the posture or gesture of the hand.
The main features of the system are the identification of signs, representation and “Text to Speech” (audio), the use of automatic learning systems and optimisation using genetic algorithms.
This same approach and technology can also be used to facilitate communication amongst persons with important physical and speech limitations.