基于集成学习支持向量机的步态识别  

GAIT RECOGNITION BASED ON SUPPORT VECTOR MACHINES AND INTEGRATIVE LEARNING

作　　者：梁竞敏[1] 

机构地区：[1]广东女子职业技术学院艺术设计与信息技术系,广东广州511450

出　　处：《计算机应用与软件》2010年第7期104-106,共3页Computer Applications and Software

基　　金：广东省科技计划项目工业攻关项目资助课题(2007B010200036)

摘　　要：提出一种基于Bagging算法和SVM的步态识别方法。首先应用背景差分法分割出运动人体轮廓,然后将人体分为多个可变区域,并通过计算获取特征向量。采用SVM分类器进行分类识别,为了提高SVM的识别率,采用Bagging算法对分类结果进行分类集成,实验结果表明,该算法取得了很好地识别性能。A novel method of gait recognition based on Bagging algorithm and support vector machines（SVM） is proposed in this paper.First,the background subtraction was used to extract body silhouette,then the dimensional silhouette was divided into several regions and the feature vectors were acquired by computation.Finally,gait classification and recognition were performed by SVM.Bagging algorithm is used to perform classification and integration against the classified outcomes for improving the recognition rate of the SVM.Experimental results demonstrate that the proposed method achieves satisfying recognition performance.

关 键 词：步态识别 BAGGING算法 支持向量机 集成学习 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]