基于肢体调整的人体姿态估计

系统仿真学报 ›› 2015, Vol. 27 ›› Issue (10): 2344-2351.

基于肢体调整的人体姿态估计

蔡鹏, 孔德慧, 尹宝才, 霍奕

北京工业大学城市交通学院多媒体与智能软件技术北京市重点实验室,北京 100124

收稿日期:2015-06-07 修回日期:2015-07-22 出版日期:2015-10-08 发布日期:2020-08-07
作者简介:蔡鹏(1977-),男,江苏句容,博士,助理研究员,博士后,研究方向为计算机图形学、计算机视觉;孔德慧,女,博士,教授,博导,研究方向为虚拟现实与图形学;尹宝才,男,博士,教授,博导,研究方向为多媒体技术、计算机图形学、视频编码。
基金资助:
国家自然科学基金(61227004; 61370120)

Pose Estimation Using Limbs Adjustment

Cai Peng, Kong Dehui, Yin Baocai, Huo Yi

Beijing Key Laboratory of Multimedia and Intelligent Software Technology, College of Metropolitan Transportation, Beijing University of Technology, Beijing 100124, China

Received:2015-06-07 Revised:2015-07-22 Online:2015-10-08 Published:2020-08-07

摘要/Abstract

摘要： 基于混合部位模型的人体姿态估计方法将人体拆分成若干由树形结构相连的模板,利用相应的全分函数进行人体姿态的估计,由于树形结构的固有缺陷而导致复杂人体姿态估计错误。为了解决这个问题,提出基于肢体调整的人体姿态估计方法。该方法提出像素离散方向的概念、融合全分函数和像素离散方向等属性的局部特征函数、以及基于关节点的局部特征进行人体姿态调整的方法。实验表明本文方法的姿态估计效果比以前的基于混合部位的关节模型的方法更好。

关键词: 混合部位, 姿态估计, 肢体调整, 像素离散方向, 全分函数

Abstract: Human body is decomposed into several local component templates connected with tree structure by articulated pose estimation with mixture-of-parts, and the corresponding full score is used to estimate body pose. Such a method fails in precisely estimating human pose due to the defects of tree models when human has the complicated pose of body. To address this problem, pose estimation using limbs adjustment with mixtures-of-parts models was proposed. And pixel discrete direction, local characteristic functions blending full score and pixel discrete direction and other attributes, and pose estimation using limbs adjustment based on local characteristic functions of articulated position were proposed, which is better than the method based on articulated pose estimation with mixtures-of-parts.

Key words: mixture-of-parts, pose estimation, limb adjustment, pixel discrete direction, full score

中图分类号:

TP391

蔡鹏, 孔德慧, 尹宝才, 霍奕. 基于肢体调整的人体姿态估计[J]. 系统仿真学报, 2015, 27(10): 2344-2351.

Cai Peng, Kong Dehui, Yin Baocai, Huo Yi. Pose Estimation Using Limbs Adjustment[J]. Journal of System Simulation, 2015, 27(10): 2344-2351.

参考文献

[1] D Marr, H Nishihara.Representation and recognition of the spatial organization of three-dimensional shapes[C]// Proceedings of the Royal Society of London. Series B, Biological Sciences, 200(1140), 1978. London, UK: Royal Society of Great Britain, 1978: 269-294.
[2] P Felzenszwalb, D Huttenlocher.Pictorial structures for object recognition[J]. IJCV (S0920-5691), 2005, 61(1): 55-79.
[3] Y Yang, D Ramanan.Articulated human detection with flexible mixtures of parts[J]. TPAMI (S0162-8828), 2013, 35(12): 2878-2890.
[4] J O’Rourke, N Badler.Model-based image analysis of human motion using constraint propagation[J]. TPAMI (S0162-8828), 1980, 2(6): 522-536.
[5] D Hogg.Model-based vision: a program to see a walking person[J]. Image and Vision Computing (S0262-8856), 1983, 1(1): 5-20.
[6] K Rohr.Towards model-based recognition of human movements in image sequences[J]. CVGIP-Image Understanding (S1077-3142), 1994, 59(1): 94-115.
[7] M. Lee, I. Cohen.Proposal maps driven mcmc for estimating human body pose in static images[C]// CVPR. Washington, DC, USA: IEEE, 2004: 334-341.
[8] L Sigal, M Black.Measure locally, reason globally: Occlusion-sensitive articulated pose estimation[C]// CVPR. New York, NY, USA: IEEE, 2006: 2041-2048.
[9] Y Wang, G Mori.Multiple tree models for occlusion and spatial constraints in human pose estimation[C]// ECCV. Marseille, France: Springer, 2008: 710-724.
[10] T Tian, S Sclaroff.Fast globally optimal 2D human detection with loopy graph models[C]// CVPR. San Francisco, CA, USA: IEEE, 2010: 81-88.
[11] V Singh, R Nevatia, C Huang.Efficient inference with multiple heterogenous part detectors for human pose estimation[C]// ECCV. Crete, Greece: Springer, 2010: 305-318.
[12] X Lan, D Huttenlocher.Beyond trees: common-factor models for 2D human pose recovery[C]// ICCV. Beijing, China: IEEE, 2005: 470-477.
[13] D Tran, D Forsyth.Improved human parsing with a full relational model[C]// ECCV. Crete, Greece: Springer, 2010: 227-240.
[14] N Dalal, B Triggs.Histograms of oriented gradients for human detection[C]// CVPR. San Diego, CA, USA: IEEE, 2005: 886-893.