432· 智能系统学报 第4卷 括平面内晃动、上下旋转、左右摇动和各种夸张表情 geles,USA,1999::187-194 等，满足了从视频中自动恢复三维人脸模型的要求： [10]AHLBERC J.Model-based coding-Extraction,coding, 同时，本文所提出的方法具有良好的计算效率，可实 and evaluation of face model parameters[D].Linkoping 时跟踪和重现人脸各种姿态和形状变化，能满足人 Ciy,Sweden:Linkoping University,2002. 机交互的要求.需要指出的是，本文方法以支持视觉 [11].VOCLER C,L Zhiguo,KANAUJA A,et al.The best of both worlds:combining 3D deformable models with active 交互为背景，与更为完备的三维形变模型相比，所获 shape models[C]l//Proceedings of the Ith IEE Interna- 取的恢复精度是有限的，也就是说，以牺牲三维恢复 tional Conference on Computer Vision.Rio de Janeiro. 精度为代价来保障算法的实时性.因此，如何在确保 Brazil,2007:H 恢复实时性前提下进一步提高恢复模型的精度是有 [12]BREGLER C.HERTZMANN A.BIERMANN H.Recove- 待进一步深入研究的问题：此外，如何将本文方法与 ing non-rigid 3D shape from image streams[C]//Proceed- 表情识别和理解等相关技术结合起来并支持人机交 ings of IEEE Conference on Computer Vision and Pattern 互过程也是普适计算和自然人机交互技术面临的重 Recognition.Hilton Head,USA.2000::690-696. 要课题， [13]BRAND M.Morphable 3D models from video[C]//Pro- ceedings of IEEE Conference on Computer Vision and Pat- 参考文献： tern Recognition.Kauai.Hawaii,USA.2001:456-463. [14].XIAOJ.CHAI J.KANADE T.A closed-form solution to [1].PORTA M.Vision-based user interfaces:methods and ap- non-rigid shape and motion recovery[C]//Proceedings of plications[J].International Journal of Human-Computer the th European Conference on Computer Vision.Prague, Studies.2002.57::27-73. Czech Republic,2004:573-587 [2]WATERS K,TERZOPOULOS D.Modeling and animating [15]ZHU Jianke,HOI S C.LYU M R.The real-time non-rigid faces using scanned data[J]Journal of Visualizaton and shape recovery via active appearance models for augmented Computer Animation,1991,2(4):123-128. reality[C]//Proceedings of the 9th European Conference [3胡永利，尹宝才，程世铨，等.创建中国人三维人脸库关 on Computer Vision.Graz,Austria,2006:186-197. 键技术研究[1.计算机研究与发展，2005,42(4)：622 16]姚伟.人脸表情识别与生成的技术研究[D].南京： 628. 南京大学，2009 HU Yongli,YIN Baocai,CHENC Shiquan,et al.Research YAO Wei.Research on facial expression recognition and on key technology in construction of a Chinese 3D face data- synthesis[D].Nanjing:Nanjing University,2009 base[J]..Journal of Computer Research and Development. 作者简介： 2005,42(4):622-628 孙正兴，男，1964年生，教授、博士 [4]IMOESLUND T B,HILTON A,KRUGER V.A survey of 生导师.中国图像图形学会计算机动画 advances in vision-based human motion capture and analysis 与数字艺术专委会常务委员，中国计算 [I].Computer Vision and Image Understanding,2006. 机学会计算机辅助设计与图形学专委 104(2)1:90-126 会委员，中国人工智能学会人工心理与 :VACCHETTI L,LEPETIT V,FUA P.Stable real-time 3D 人工情感专委会委员，江苏省微型电脑 tracking using online and offline information [J].IEEE 应用协会副理事长兼多媒体技术专委会主任，江苏省计算机 Transactions on Pattern Analysis and Machine Intelgence, 学会计算机辅助设计与图形学专委会主任.国家自然科学 2004,26(10)1:1385-1391. 基金委创新研究群体(007年度)骨干成员.主要研究方 [6]|LIU Z.ZHANG Z,JACOBS C,et al.Rapid modeling of 向为多媒体计算、计算机视觉和环境智能.获省部级科技进 animated faces from video[J].The Journal of Visualization 步三等奖3次.已在国内外重要学术刊物上发表学术论文 and Computer Animation,2001,12(4):227-240. 90余篇，主编教材3部、译著1部 [7]COOTES TF.TAYLOR CJ.COOPER D H.et al.Active shape models-their training and application[J].Computer 姚伟，男，1984年生，硕士研究 Vision and Image Understanding,1995,61(1)):38-59. 生，主要研究方向为计算机视觉与智能 8COOTES T F,EDWARDS C,TAYLOR C J.Active ap- 人机交互， pearance models[J].IEEE Transactions on Pattern Analy- sis and Machine Intelligence,2001,23(6)':681-685. BLANZV,VETTER T.A morphable model for the synthe- sis of 3D-face[C]//Proceelings of ACM International Con- ference on Computer Graphics(SIGGRAPH 99).Los An-16] 姚 伟.人脸表情识别与生成的技术研究[D] .南京: 南京大学,2009. geles,USA,1999: 187-194. 作者简介: [5] VACCHETTI L,LEPETIT V,FUA P. Stable real-time 3D tracking using online and offline information [J]. IEEE Transactions on Pattern Analysis and Machine Intelgence, 2004,26(10) : 1385-1391. [14] XIAOJ,CHAI J,KANADE T. A closed-form solution to non-rigid shape and motion recovery[ C] //Proceedings of the th European Conference on Computer Vision. Prague, Czech Republic,2004: 573-587. [9] BLANZV,VETTER T. A morphable model for the synthe￾sis of 3D-face[ C] //Proceelings of ACM International Con￾ference on Computer Graphics(SIGGRAPH 99). Los An- [3] 胡永利,尹宝才,程世铨,等.创建中国人三维人脸库关 键技术研究[J] .计算机研究与发展,2005,42(4) : 622- 628. [6] LIU Z,ZHANG Z,JACOBS C,et al. Rapid modeling of animated faces from video[J] . The Journal of Visualization and Computer Animation,2001,12(4) :227-240. 姚 伟,男,1984年生,硕士研究 生,主要研究方向为计算机视觉与智能 人机交互. [15]ZHU Jianke,HOI S C,LYU M R. The real-time non-rigid shape recovery via active appearance models for augmented reality[C]//Proceedings of the 9th European Conference on Computer Vision. Graz,Austria,2006: 186-197. [12] BREGLER C,HERTZMANN A,BIERMANN H.Recove￾ing non-rigid 3D shape from image streams[ C] //Proceed￾ings of IEEE Conference on Computer Vision and Pattern Recognition. Hilton Head,USA,2000: 690-696. 孙正兴,男,1964年生,教授、博士 [1] PORTA M. Vision-based user interfaces: methods and ap￾plications[J] 生导师.中国图像图形学会计算机动画 与数字艺术专委会常务委员,中国计算 机学会计算机辅助设计与图形学专委 会委员,中国人工智能学会人工心理与 人工情感专委会委员,江苏省微型电脑 应用协会副理事长兼多媒体技术专委会主任,江苏省计算机 学会计算机辅助设计与图形学专委会主任.国家自然科学 基金委创新研究群体(2007年度) 骨干成员.主要研究方 向为多媒体计算、计算机视觉和环境智能.获省部级科技进 步三等奖3次.已在国内外重要学术刊物上发表学术论文 90余篇,主编教材3部、译著1部. . International Journal of Human-Computer Studies,2002,57: 27-73. [8] COOTES T F,EDWARDS C,TAYLOR C J. Active ap￾pearance models[J]. IEEE Transactions on Pattern Analy￾sis and Machine Intelligence,2001,23(6) :681-685. [13] BRAND M. Morphable 3D models from video[C] //Pro￾ceedings of IEEE Conference on Computer Vision and Pat- 参考文献: tern Recognition. Kauai,Hawaii,USA,2001: 456-463. YAO Wei. Research on facial expression recognition and synthesis[D].Nanjing: Nanjing University,2009. 智 能 系 统 学 报 第4卷 [10] AHLBERC J. Model-based coding-Extraction,coding, and evaluation of face model parameters[D] . Linkoping Ciy,Sweden: Linkoping University,2002. [2] WATERS K,TERZOPOULOS D. Modeling and animating faces using scanned data[J].Journal of Visualizaton and Computer Animation,1991,2(4) :123-128. [4] MOESLUND T B,HILTON A,KRUGER V. A survey of advances in vision-based human motion capture and analysis [J] . Computer Vision and Image Understanding,2006, 104(2) : 90-126. [7] COOTES TF,TAYLOR CJ,COOPER D H,et al. Active shape models-their training and application[J]. Computer Vision and Image Understanding,1995,61(1) :38-59. HU Yongli,YIN Baocai,CHENC Shiquan,et al. Research on key technology in construction of a Chinese 3D face data￾base[J] . Journal of Computer Research and Development, 2005,42(4) : 622-628. 432· 括平面内晃动、上下旋转、左右摇动和各种夸张表情 等,满足了从视频中自动恢复三维人脸模型的要求; 同时,本文所提出的方法具有良好的计算效率,可实 时跟踪和重现人脸各种姿态和形状变化,能满足人 机交互的要求.需要指出的是,本文方法以支持视觉 交互为背景,与更为完备的三维形变模型相比,所获 取的恢复精度是有限的,也就是说,以牺牲三维恢复 精度为代价来保障算法的实时性.因此,如何在确保 恢复实时性前提下进一步提高恢复模型的精度是有 待进一步深入研究的问题;此外,如何将本文方法与 表情识别和理解等相关技术结合起来并支持人机交 互过程也是普适计算和自然人机交互技术面临的重 要课题. [11] VOCLER C,L Zhiguo,KANAUJA A,et al. The best of both worlds: combining 3D deformable models with active shape models[ C] //Proceedings of the 1th IEE Interna￾tional Conference on Computer Vision. Rio de Janeiro, Brazil,2007: 1-7