结合无符号Laplace谱特征的触觉步态识别算法-计算机工程与应用.PDFVIP

214 2016 ，52（1） Computer Engineering and Applications 计算机工程与应用 结合无符号Laplace 谱特征的触觉步态识别算法 1，2 1，2 鲍文霞 ，梁 栋 BAO Wenxia1，2 , LIANG Dong1，2 1.安徽大学 计算机智能与信号处理教育部重点实验室，合肥 230039 2.安徽大学 电子信息工程学院，合肥 230601 1.Key Lab of Intelligent Computing Signal Processing, Ministry of Education, Anhui University, Hefei 230039, China 2.School of Electronics and Information Engineering, Anhui University, Hefei 230601, China BAO Wenxia, LIANG Dong. Algorithm on tactile gait recognition combined with signless Laplace spectrum fea- ture. Computer Engineering and Applications, 2016, 52 （1）：214-218. Abstract: Aiming at the lack of tactile gait recognition only by using the pressure distribution feature, an algorithm on dynamic tactile gait recognition combined with signless Laplace spectrum is put forward. Tactile gait data under constant, fast or slow speed is collected. Then plantar pressure images are obtained, which can be divided into regions according to the plantar anatomy structure. The structure graph is constructed with regions of plantar pressure image for the nodes. And it is expressed in signless Laplace matrix. The spectrum feature is gained by using singular value decomposition to the matrix. The one-against-one support vector machine classification method is employed. Then the classifiers are designed in different speed. So gait recognition is realized. Experimental results show the recognition accuracy of the algorithm is high in different speed. Key words: signless Laplace spectrum; plantar pressure image; tactile gait recognition; Support Vector Machine（SVM ） 摘 要：针对单纯利用压力点分布特征进行触觉步态识别的不足，提出了一种结合无符号Laplace 谱特征的动态触觉步 态识别算法。利用足底压力数字化场地采集常速、快速和慢速三种情况下的触觉步态数据，生成足底压力分布图像，并 根据足底解剖学的结构划分区域；以足底压力图像各区域为节点构造结构图，并采用无符号Laplace 矩阵表示；通过对 该矩阵进行奇异值分解（Singular Value Decomposition，SVD）获取谱特征，并结合形状特征得到触觉步态特征；选择 “一对一”的支持向量机（Support Vector Machine，SVM）多分类方法，按照人在行走过程中不同的速度分别构造分类 器，从而实现动态触觉步态的识别。实验结果表明该识别算法对不