A Fundamental Study of an Animation Presentation System for JSL Signs
Mina Terauchi*, Yuji Nagashima**, Kazuyuki Kanda***,
Takateru Nishimura**, Hideyo Nagashima** and Genichi Ohwa*
*Polytechnic University, 4-1-1, Hashimotodai, Sagamihara-shi, Kanagawa, 229 Japan
**Kogakuin University, 2665-1 Nakanomachi, Hachioji-shi, Tokyo, 192 Japan
***Chukyo University, 10ITokodate, Kaizucho, Toyota-shi, 470-03 Japan
We describe a system of presenting the animation of the Japanese Signs by a Key-Frame Method as part of a supporting system for hearing-impaired persons. According to this system, JSL Signs are extracted from Japanese sentences whose elemental forms have been analyzed, and sign language corresponding to each sign is presented in image by using the angle displacement of each joint as a parameter. As a result of obtaining animated sign presentations by using several sample sentences, it was possible to recognize signs. However, since presented images are skelton models, it will be required in the future to examine a more effective image presentation system for the cases in which more realistic models are presented. In addition, in order to solve various problems associated with Key-Frame Method, we also examined an animated presentation system with the concept of Object-Oriented introduced. Taking note of the fact that a human is a multi-joint structure, we define a descriptive form in which the movement of each rigid body and the positional relationship between rigid bodies are treated as knowledge. By so doing, it now becomes possible to provides more flexible descrive forms and motion expressions.