基于发音运动特征的舌部肌肉建模及仿真

Modeling and simulation of tongue muscle based on pronunciation characteristics

人舌结构较为复杂,运动灵活,在虚拟人物发音及表情运动过程中起着关键的作用,但现有舌部模型结构较单一,实现的虚拟运动相对简单,描述能力不足,且需要大量数据驱动。针对以上问题,通过对舌部解剖结构的研究,提出一种基于肌肉运动特征的舌部模型及控制方法。将舌部内外肌群按其运动及发音功能进行分类,仅需获得多条肌肉不同的收缩量,即可利用各自的控制函数确定出参数值,进而实现模型前升高、后升高和谷形等多种常见运动。实验结果表明,该方法能够较为灵活便捷地控制虚拟舌体,仿真出符合生理特征的多种舌部运动。

The human tongue structure is complex, and its movement is flexible. It plays a key role in the process of pronunciation and expression of movement in virtual characters. But the structures of the existing tongue models are sole, and the virtual movements are relatively simple. Besides, those models are lack of describing ability, and require a lot of data-driven. To solve these above problems, through the study of anatomical structure of the tongue, a tongue muscle model is put forward which is based on the characteristics of muscle movements. According to the motions and pronunciation function, the tongue muscles groups, including the inside and outside muscles, are divided into some categories. It only gets the different shrinkages about those muscles, and takes advantage of each control function to determine the parameter values. Thus, variety of common movements can be achieved, such as front rise, back rise and valley-shaped. The experimental results show that this method can be more flexible and convenient to control a virtual tongue, and simulate a variety of physiological characteristics movements of the tongue.