基于动作捕获的三维动画人物表情变形自动控制  

Automatic Control of Expression Deformation of 3D Animation Character Based on Action Capture

作　　者：李作山 LI Zuo-shan(School of Information Engineering,Suihua University,Suihua,152000,Heilongjiang)

机构地区：[1]绥化学院信息工程学院,黑龙江绥化152000

出　　处：《蚌埠学院学报》2021年第2期61-64,共4页Journal of Bengbu University

基　　金：黑龙江省艺术科学规划项目(2019D014)。

摘　　要：为提高三维动画人物表情变形控制的精准度与收敛性,提出基于动作捕获的三维动画人物表情变形自动控制方法。建立三维动画人物表情图像的视觉信息采集模型,采用相关特征点检测和匹配方法检测表情变形特征,提取关联分布特征点。通过模糊信息融合和空间像素聚类分析方法,检测表情变形状态。结合显著性特征提取的方法,捕获变形特征动作,实现三维动画人物表情变形的自动控制。仿真结果表明,该方法提取的变形特征与样本表情一致,控制过程的收敛性较好。In order to improve the accuracy and convergence of 3D animation character facial expression deformation control,an automatic control method based on motion capture was proposed.The visual information acquisition model of 3D animated character facial expressions was established,and the facial expressions deformation features were detected by using relevant feature point detection and matching method,and the associated distribution feature points were extracted.Fuzzy information fusion and spatial pixel clustering analysis were used to detect the deformation of facial expressions.Combined with the method of significance feature extraction,the deformation feature movement was captured to realize the automatic control of the deformation of 3D animated characters’expressions.The simulation results showed that the deformation features extracted by this method are consistent with the sample expressions,and the convergence of the control process is good.

关 键 词：动作捕获 三维动画 人物表情 变形 自动控制 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]