A Framework for a Sign Language Interfacing System

Sign languages have been proven to be natural languages, as capable of expressing human thoughts and emotions as traditional (spoken) languages are. The distinct visual and spatial nature of sign languages makes it difficult to develop an interfacing system as a communication medium platform for sign language users. This dissertation targets this problem by presenting some explorations in the areas of computer graphics, interface design, and human-computer interactions with emphasis on software development and implementation. First, a virtual human body has been constructed as a set of functional body components with the consideration of the biomechanics of the human body. Then based on the virtual body, virtual gestures are created by controlling the movements of the functional components, with substantial concentration on the creation of natural hand configurations and the application of body joint motion constraints. Finally, from the virtual gestures, sign language linguistic parts can be constructed by using the Movement-Hold Model. An ease-of-use graphical user interface is designed and implemented that wraps up of all the operations of gesture generation and editing, gesture database management, and creation, editing, storage, and retrieval of sign language linguistic parts. In this dissertation, we propose a sign language interfacing system with verification of preliminary results that differs from other computer-related sign language systems in its (1) flexibility of creating any sign language linguistic parts (from phonemes to sentences), (2) “openness” of application to any sign languages, and (3) capability of being used as a sign language “editor” like a word processor for spoken languages.