基于手势数据分析的地方手语动画合成方法研究

摘要

为了在现有的手势数据基础上合成新的地方手语手势,采取对这些手势数据进行运动相似性分段和动态聚类自动获取具有时序特征的基本手形类数据以及结合手臂状态数据的方法,最后生成新的地方手语手势动画数据.这不仅节省了重新采集数据需要的昂贵设备和大量工作,而且也能达到手势表达准确性的目的.%In order to synthesize dialect sign language gesture based on existing gesture data, the paper used the methods that segmented gesture data according to its motion similarity and dynamically clustered them to get basic hand gesture cluster data with time sequence feature automatically. Then combined arm condition data to synthesize new dialect sign gesture animation data. The research not only saves expensive device and a great deal of work, but also achieves gesture expression' s accuracy.