摘要
脸部建模是特效电影和电脑游戏中展现现代视觉效果的关键。本文针对三维人脸模型的动态建模,提出一种基于特征提取的建模方法。对于捕获的人脸图像,首先定位其人脸区域,然后提取人脸区域的人脸特征点,同时根据提取得到的人脸特征点对标准三维人脸模型进行形变处理,最后得到一个实时的三维人脸动画系统。实验结果表明,该方法针对三维人脸建模能够实时稳定地完成相应表情的人脸建模,具有较高的实时性、准确性。
Face modeling is the key step to modern visuals in special effects movies and computer games.In this paper,aiming the dynamic modeling of 3D face model,a modeling method based on feature extraction is proposed.For the captured face image,locate the face region firstly,then extract the face feature points of the face region,and deform the standard 3D face model according to the extracted face feature points,and finally obtain a real-time three-dimensional face animation system.The experimental results show that the proposed method can accurately complete the face modeling of the corresponding expression in real time for 3D face modeling,which has high real-time and accuracy.
作者
杨金秋
童立靖
付孝琴
YANG Jin-qiu;TONG Li-jing;FU Xiao-qin(School of Computer Science and Technology,North China University of Technology,Beijing 100144)
出处
《数字技术与应用》
2018年第9期34-36,共3页
Digital Technology & Application
基金
2018年北方工业大学大学生科技活动项目资助
关键词
人脸建模
特征提取
径向基函数网络
动画系统
face modeling
feature extraction
radial basis function network
animation system