第1期 肖锋等：基于区域分割和蒙特卡洛采样的静态图片人体姿态估计 43. 5结束语 [10]HU Zhilan,WANG Guijin,LIN Xinggang,et al.Recov- ery of upper body poses in static images based on joints de- 本文提出了一种新的姿态估计的算法，与现有 tection[J].Pattern Recognition Letters,2009,30(5): 方法相比，存在如下改进：在采用特征方面，考虑了 503-512. 人体的对称性以及皮肤、各部分颜色的均衡性等更 [11]FERRARI V.MARIN-JIMENEZ M.ZISSERMAN A.Pro-- gressive search space reduction for human pose estimation 适合描述人体的特征：在推理算法方面，采用蒙特卡 [C]/IEE Computer Society Conference on Computer Vi- 洛算法提高了计算速度；同时采用了非树状的约束 sion and Pattern Recognition.Anchorage,USA,2008:1-8. 关系，使得精度得到了提高.在后续工作中，将研究 [12]ZHANG Xiaogin,LI Changcheng,TONG Xiaofeng.et al. 如何改进特征描述，以期提出更好的特征描述方法， Efficient human pose estimation via parsing a tree structure 提高人体姿态检测的性能 based human model[C]//IEEE International Conference 参考文献： on Computer Vision.Kyoto,Japan,2009:1349-1356. [3]JBERGTHOLDT M,JAPPES J,SCHMIDT S,et al.A. [1]EICHNER M,FERRARI V.Better appearance models for study of parts-based object class detection using complete pictorial structures[C]//Proceedings of British Machine Vi- graphs[J].International Journal of Computer Vision,, sion Conference.London,UK,2009::6-17. 2010,87(1/2):93-117. [2)杨金锋付惆宇，谭铁牛，等一种新型的基于内容的图 [14]CHEN Qiang,ZHENG Enliang,LIU Yuncai.Pose estima- 像识别与过滤方法[J]:.通信学报，2004,25(7)：98 tion based on human detection and segmentationJ].Sci- 106. ence in China Series F:Information Sciences,2009,52 YANG Jinfeng,FU Zhouyu,TAN Tieniu,et al.A novel al- ②)：244-251. gorithm for content-based image recognition and filtering 15]]BOURDEV L.MALIK J.Poselets:body part detectors .Journal of China Institute of Communications,2004, trained using 3D human pose annotation[C]//IEEE Inter-. 25(7):93-106 national Conference on Computer Vision.Kyoto,Japan,. 3 HU Weiming,WU Ou,CHEN Zhouyao,et al.Recognition 2009::1365-1374. of pornographic web pages by classifying texts and images [16]RAMANAN D.Learning to parse images of articulated [IEEE Transactions on Pattern Analysis and Machine bodies[C]//Proceedings of the Twentieth Annual Confer-- Intelligence,2007,29(6):1019-1034. ence on Neural Information Processing Systems.Vancou-- 4]FELZENSZWALB P.HUTTENLOCHER D.Pictorial struc- ver,Canada,2006::1129-l136. ures for object recognition[J].International Journal of [17]JONES M,REHG J.Statistical color models with applica- Computer Vision,2005,61(1)):55-79. tion to skin detection[J].International Joumal of Comput- ]MORI G,REN Xiaofeng,EFROS A,et al.Recovering hu- er Vision,2002,46(1):81-96. man body configurations:combining segmentation and rec- 18]YEDIDIA J,FREEMAN W.WEISS Y.Understanding be- ognition[C]//IEEE Computer Society Conference on Com- lief propagation and its generalizations[J].Exploring Arti- puter Vision and Pattern Recognition.Washington,DC, ficial Intelligence in the New Millennium,2002,8(1)): USA,2004:326-333. 236-239」 6 IREN Xiaofeng,BERG A,MALIK J.Recovering human 作者简介： body configurations using pairwise constraints between parts 肖锋，男，1982年生，博士研究生 [C]//EEE Intemational Conference on Computer Vision. 主要研究方向为姿态估计、图像处理、 Beijing,China,2005:824-831. 图片分类 [7]ROBERTS T,MCKENNA S,RICKETTS I.Human pose estimation using partial configurations and probabilistic re- gions[J].International Journal of Computer Vision,2007, 73(3)1:285-306. LEE M,COHEN I.A model-based approach for estimating 周杰，男，1968年生，教授、博士生 human 3D poses in static images[J].IEEE Transactions on 导师，主要研究方向为模式识别、计算】 Pattern Analysis and Machine Intelligence,2006,28(6)): 机视觉、多媒体信息处理、信息服务科 905-916. 学与技术.承担或完成国家自然科学基 [9]HUA Gang,YANG Minghsuan,WU Ying.Learning to esti 金、“863“计划项目等10余项，中国航 mate human pose with data driven belief propagation[C]// 海学会科学技术奖二等奖等多项奖励， EEE Computer Society Conference on Computer Vision and 在国外重要学术期刊上发表学术论文20余篇： Pattem Recognition.San Diego,USA,2005::747-754[8] LEE M,COHEN I. A model-based approach for estimating human 3D poses in static images[J] . IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(6) : 905-916. [12]ZHANG Xiaoqin,LI Changcheng,TONG Xiaofeng,et al. Efficient human pose estimation via parsing a tree structure based human model[C] //IEEE International Conference on Computer Vision. Kyoto,Japan,2009: 1349-1356. 肖锋,等: 周杰,男,1968年生,教授、博士生 导师,主要研究方向为模式识别、计算 机视觉、多媒体信息处理、信息服务科 学与技术.承担或完成国家自然科学基 金、"863"计划项目等10余项,中国航 海学会科学技术奖二等奖等多项奖励, 在国外重要学术期刊上发表学术论文20余篇. 基于区域分割和蒙特卡洛采样的静态图片人体姿态估计 [16] RAMANAN D. Learning to parse images of articulated bodies[C]//Proceedings of the Twentieth Annual Confer￾ence on Neural Information Processing Systems. Vancou￾ver,Canada,2006: 1129-1136. [1] EICHNER M,FERRARI V. Better appearance models for pictorial structures[ C]//Proceedings of British Machine Vi￾sion Conference. London,UK,2009: 6-17. YANG Jinfeng,FU Zhouyu,TAN Tieniu,et al. A novel al￾gorithm for content-based image recognition and filtering [J].Journal of China Institute of Communications,2004, 25(7) : 93-106. [13] BERGTHOLDT M,JAPPES J,SCHMIDT S,et al. A study of parts-based object class detection using complete graphs[ J]. International Journal of Computer Vision, 2010,87(1/2) : 93-117. 6 ] REN Xiaofeng,BERG A,MALIK J. Recovering human body configurations using pairwise constraints between parts [C] //IEEE International Conference on Computer Vision. Beijing,China,2005: 824-831. [10] HU Zhilan,WANG Guijin,LIN Xinggang,et al. Recov￾ery of upper body poses in static images based on joints de￾tection[ J]. Pattern Recognition Letters,2009,30(5) : 503-512. [14] CHEN Qiang,ZHENG Enliang,LIU Yuncai. Pose estima￾tion based on human detection and segmentation[J] . Sci￾ence in China Series F: Information Sciences,2009,52 (2) : 244-251. 4 ] FELZENSZWALB P,HUTTENLOCHER D. Pictorial struc￾ures for object recognition[J]. International Journal of Computer Vision,2005,61(1) :55-79. 5结束语 本文提出了一种新的姿态估计的算法,与现有 方法相比,存在如下改进: 在采用特征方面,考虑了 人体的对称性以及皮肤、各部分颜色的均衡性等更 适合描述人体的特征;在推理算法方面,采用蒙特卡 洛算法提高了计算速度;同时采用了非树状的约束 关系,使得精度得到了提高.在后续工作中,将研究 如何改进特征描述,以期提出更好的特征描述方法, 提高人体姿态检测的性能. [11] FERRARI V,MARIN-JIMENEZ M,ZISSERMAN A. Pro￾gressive search space reduction for human pose estimation [C] //IEE Computer Society Conference on Computer Vi￾sion and Pattern Recognition. Anchorage,USA,2008: 1-8. 3 ] HU Weiming,WU Ou,CHEN Zhouyao,et al. Recognition of pornographic web pages by classifying texts and images [J] . IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(6) : 1019-1034. [9] HUA Gang,YANG Minghsuan,WU Ying. Learning to esti mate human pose with data driven belief propagation[C]// EEE Computer Society Conference on Computer Vision and Pattem Recognition. San Diego,USA,2005: 747-754. 参考文献: 43. [18] YEDIDIA J,FREEMAN W,WEISS Y. Understanding be￾lief propagation and its generalizations[J]. Exploring Arti￾ficial Intelligence in the New Millennium,2002,8(1) : 236-239 15] BOURDEV L,MALIK J. Poselets: body part detectors trained using 3D human pose annotation[ C] //IEEE Inter￾national Conference on Computer Vision. Kyoto,Japan, 2009: 1365-1374. 肖锋,男,1982年生,博士研究生, 主要研究方向为姿态估计、图像处理、 图片分类. 2]杨金锋,付周宇,谭铁牛,等.一种新型的基于内容的图 像识别与过滤方法[J] .通信学报,2004,25(7) : 93- 106. 第1期 [5] MORI G,REN Xiaofeng,EFROS A,et al. Recovering hu￾man body configurations: combining segmentation and rec￾ognition[ C]//IEEE Computer Society Conference on Com￾puter Vision and Pattern Recognition. Washington,DC, USA,2004: 326-333. [7] ROBERTS T,MCKENNA S,RICKETTS I. Human pose estimation using partial configurations and probabilistic re￾gions[J]. International Journal of Computer Vision,2007, 73(3) :285-306. [17]JONES M,REHG J. Statistical color models with applica￾tion to skin detection[J]. International Journal of Comput￾er Vision,2002,46(1) : 81-96. 作者简介: