融合RGB-D多模特征和半径边缘约束超限学习的动作识别  

Action Recognition Fused with RGB-D Multi-modal Features and Radius-Margin Bounded Extreme Learning

作　　者：刘天亮[1] 陈克虎 戴修斌[1] 罗杰波 LIU Tianliang1;CHEN Kehu1;DAI Xiubin1;LUO Jiebo2

机构地区：[1]南京邮电大学江苏省图像处理与图像通信重点实验室,南京210003 [2]罗彻斯特大学计算机科学系,罗彻斯特14627

出　　处：《模式识别与人工智能》2018年第10期958-964,共7页Pattern Recognition and Artificial Intelligence

基　　金：国家自然科学基金项目(No.61001152;31200747;61071091;61071166;61172118);江苏省自然科学基金项目(No.BK2012437);南京邮电大学校级科研基金项目(No.NY214037);国家留学基金项目资助~~

摘　　要：针对单模态特征鉴别行为动作类别的能力有限问题,提出基于RGB-D视频中多模态视觉特征融合和实例化多重核超限学习(Exemplars-MKL-ELM)的动作分类方法.首先,利用骨架表面拟合和密集轨迹提取稳健的密集运动姿态特征,以稠密点云法平面感知人体3维几何的稀疏化有向主成分直方图特征,提取外观纹理嵌入身体节点空-时邻域的三维梯度直方图特征.然后,采用半径边缘约束多重核超限学习机融合多模态视觉特征,并利用对比数据法挖掘每个行为类别的代表性实例集合.最后,每个样本结合融合视觉特征和即得实例集合,采用Exemplars-MKL-ELM模型和贪婪预测思想分层分类识别行为.实验表明,文中方法在分类准确度和计算效率上都较优.An exemplars multiple kernel learning-extreme learning machine （MKL-ELM） based human action recognition approach with muhi-modal visual feature fusion from RGB-D videos is proposed to solve the problem of single modal visual feature with the limited discrimination ability for all categories of human actions. Firstly, the robust and dense moving pose features with human skeleton surface fitting and dense trajectories from human motion are extracted. The sparse histogram of oriented principal component （SHOPC） features of 3D body geometry with the normal plane of dense point clouds is perceived and the histogram of 3D gradient orientation （HOG3D） features embedded with human appearance textures on spatial temporal neighbor of body nodes in the given videos is extracted. The modified MKL-ELM with radius-margin bound is exploited to fuse the given multi-modal visual features. Then, the set of the representative exemplars for each human action is mined with the contrast data technique. Finally, each sample is hierarchically classified by the designed exemplars-MKL-ELM model with greedy prediction strategy to recognize the human actions with the fused features and the givenexemplars. The experiments show that compared with the traditional methods, the proposed action recognition method has significant advantages with high classification accuracy and computational efficiency.

关 键 词：多模态特征 多核学习 超限学习机 RGB-D视频 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]