Automatic transformation of Russian manual-alphabet gestures into textual form

This paper considers the task of sign-language interpretation for gestures that are used in the Russian Manual Alphabet (RMA) for the deaf and dumb. A software-hardware system is suggested that permits static and dynamic gestures to be transformed into a text on-line; the conducted experiments show that the system currently provides a fairly reliable recognition of all static and some dynamic RMA gestures. Methods to improve the quality of the dynamic gesture analysis are planned. © 2014, Allerton Press, Inc.