融合真实照片创建三维模型的影视动画制作研究  

Research on Film and TV Animation Production by Integrating Real Photos to Create 3D Models

作　　者：吴志诚[1] 陶玲凤[1] WU Zhi-cheng;TAO Ling-feng(School of Arts and Media, Wuhu Institute of Technology, Wuhu 241000,China)

机构地区：[1]芜湖职业技术学院艺术传媒学院,安徽芜湖241000

出　　处：《信阳农林学院学报》2018年第4期125-128,132,共5页Journal of Xinyang Agriculture and Forestry University

摘　　要：通过对真实照片的多角度动态扫描和识别来构建照片中人和物的三维建模技术广泛应用于三维游戏制作、影视动画设计、影像医学等领域,该技术可以为用户提供逼真的三维真实体验。基于上述背景,提出了一种基于分区识别的一般人脸特征提取算法,可以利用提取的有限人脸核心特征点快速构建人脸三维逼真模型。为了验证所提算法的有效性,基于OpenGL三维开源插件在VS2008环境下开发了验证环境并对算法进行了实际验证,验证结果表明基于分区识别的一般人脸特征提取算法可以在较短的时间内有效重建人脸三维逼真模型,面部表情、面部纹理信息等关键点符合真实人脸特征。The3D modeling technology of building people and objects in photos by multi angle dynamic scanning and recognition of real photos is widely used in3D game making,film and television animati on design,image medicine and so on.This technology can provide realistic three-dimensional real experience for users.Based on the above background,a general face feature extraction algorithm based on partition recognition is proposed in this paper,which can make use of the core feature points extracted from the limited face to quickly construct the3D realistic model of face.In order to verify the effectiveness of the proposed algorithm,this paper has developed a verification environment based on the OpenGL three-dimensional open source plug-in in the VS2008environment and verified the algorithm.The results show that the proposed algorithm based on partition recognition in this paper can effectively reconstruct the3D realistic model of face in a short time.The key points of facial expression and facial texture information conform to the real facial features.

关 键 词：真实照片 三维人脸建模 特征提取 OPENGL 影视动画 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]