摘要
提出一种改进的基于地面反作用力的步态识别方法.该方法通过由三维测力台构建的步态通道获取步行时足底受到的三方向地面反作用力,并采用小波包分解提取时频域特征,利用模糊C-均值聚类算法从中挑选出最具分类能力的特征子集,最后在训练样本上用支持向量机训练分类器,并在测试集上进行步态识别.为提高识别率,对样本进行拆分和波形对齐操作,并设计多分类器以降低步行速度变化对识别准确率的影响.在103人的步态数据库上的测试结果表明,该方法即使在训练样本较少的情况下也可以得到较高的识别率.
An improved gait recognition approach based on ground reaction force (GRF) is proposed. 3-directional GRF are acquired by 3-dimensional force plate while a person is walking through the gait walkway. Wavelet packet (WP) decomposition is used to extract features in time-frequency domain, and optimal feature subset is selected using a fuzzy c-means (FCM) clustering algorithm. Support vector machine (SVM) classifier is trained on training-set, and then gait recognition is implemented by SVM on testingset. In order to improve the recognition accuracy, waveform alignment and re-sampling approach are utilized. Multiple classifiers are designed to reduce the negative influence of changes in walking speed. The approach is tested on a gait database collected from 103 subjects. Comparative results demonstrate that high recognition accuracy can be reached even in fewer training-samples.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2011年第3期353-359,共7页
Pattern Recognition and Artificial Intelligence
基金
中国科学院合肥物质科学研究院院长基金(基于步态触觉信息的生物特征识别方法研究
0722B11141)
关键词
步态识别
地面反作用力
小波包分解
模糊C-均值
多分类器
Gait Recognition, Ground Reaction Force, Wavelet Packet Decomposition, Fuzzy C-Means, Multiple Classifiers
作者简介
林尔东,男,1986年生,硕士,主要研究方向为模式识别.E-mail:lined8963@gmail.com.
姚志明,男,1983年生,博士,主要研究方向为运动生物力学与模式识别.
孙怡宁,男,1963年生,教授,博士生导师,主要研究方向为传感技术、运动生物力学、系统集成.E-mail:ynsun@iim.ac.cn.