一种步态触觉特征识别算法  

A recognition algorithm based on gait touching feature

作　　者：王琨[1] 王年[1] 鲍文霞[1] 苏亮亮[1] 

机构地区：[1]安徽大学计算智能与信号处理教育部重点实验室,安徽合肥230039

出　　处：《安徽大学学报（自然科学版）》2013年第5期32-36,共5页Journal of Anhui University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61172127);安徽省自然科学基金资助项目(1208085MF93);高等学校博士学科点科研基金资助项目(20113401110006);安徽省高校优秀青年人才基金资助项目(2012SQRL017ZD);安徽大学"211工程"学术创新团队基金资助项目(KJTD007A)

摘　　要：鉴于不同人足底的尺寸和压力分布的不同,提出了一种基于压力中心点距离的步态触觉特征识别算法,通过对足底脚趾区、前脚掌区和后脚掌区的划分及其压力中心点的标记,计算不同区域压力中心点间的欧氏距离以获取相应足底压力的特征表示,从而实现人的身份识别.论文基于该特征结合支持向量机(support vector machine,SVM)分类器对真实数据进行了人的身份识别,实验结果验证了该算法的有效性.Due to the different sizes and dissimilar pressure distributions of human beings＇ feet, a new algorithm based on distance of center pressure was proposed to distinguish different people＇s footprints. With the division of range of toe area, fore sole area and heel area of human＇ foot, as well as marking the global center pressure point, the distances of the global center to the centers of three different areas could be obtained. Thus the feature of certain foot would be acquired. Finally, with this kind of feature, identity recognition would be realized. This paper dealed with sole gait recognition analysis of real data by combining the feature with support vector machine, and the feasiblility and effective of the result in identity recognition were verified in this paper.

关 键 词：步态 特征提别 识别 支持向量机 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]