计算机技术与发展2018，Vol.28Issue(1)：17-22,6.DOI:10.3969/j.issn.1673-629X.2018.01.004
基于融合特征的群体行为识别
Recognition of Human Group Action Based on Fusion Features
摘要
Abstract
Based on the study of feature extraction of group behavior,we propose a group action recognition method based on fusion move-ment characteristics and appearance characteristics. In order to effectively describe the identification information,the trajectories of each pe-destrians are calculated by covariance tracking to gain the nodes of crowd network. The Granger causality test is used to estimate the relation-ship between pedestrians. Based on these causations,two types of complex network are generated which are pair-complex network and group-complex network,and the appearance information is also included in the description of group actions. Finally,we adopt the support vector machine ( SVM) based on the improved glowworm swarm optimization ( GSO) to recognize human group action. Experiment shows that the proposed method can express and recognize group action effectively.关键词
群体行为识别/特征融合/Granger因果/支持向量机Key words
group action recognition/feature fusion/Granger causality/SVM分类
信息技术与安全科学引用本文复制引用
谭程午,夏利民,王嘉..基于融合特征的群体行为识别[J].计算机技术与发展,2018,28(1):17-22,6.基金项目
国家自然科学基金资助项目(50808025) （50808025）
