摘要
本文研究真实感虚拟手语主持人。使用复杂度相对较低的漫反射环绕算法来近似实现子面散射效果,以满足系统实时渲染效果。针对手语主持人手势动作不清晰情况,提出了一种勾勒手指轮廓边缘的算法,增强了手指的立体感。利用蒙皮技术实现的唇部模型可以在实时要求下模拟出更为细致的唇动变化。利用手势时长来指导多种模式间的协同,取得了良好的实用效果。
This paper focuses on vivid virtual sign language comperes. To meet the real-time render effect of the display system,a less complexity method to imitate SubSurface Scattering which is called diffuse surround algorithm is introduced. An edge emphasis algorithm aiming at fingers' outline is also proposed in this paper to enhance third dimension. Skin mesh technology is employed to produce much more detailed changes in lip movement. Using gesture duration guides the multi-model behavior synchronizing works well.
出处
《微计算机信息》
2010年第31期217-219,共3页
Control & Automation
关键词
虚拟人
真实感
可懂度
电视手语
avatar
reality
intelligibility
TV sign language