欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出一种可用于门禁和考勤系统的静态触觉步态识别方法.首先,对压力测试板采集的裸脚静止足底压力图像进行划分区域、去噪和筛选等预处理以消除噪声点对特征提取的影响；其次,根据足底不同区域最大压力点和压力中心点的坐标及压力值提取拉普拉斯谱特征,同时结合足底形状及压力比值特征以提高算法的鲁棒性；最后,利用一对一的支持向量机（SVM）多分类方法在50人左、右脚的静态步态数据上训练分类器并进行分类识别,结果表明该算法受外界干扰小,其平均识别率达96％.
英文摘要：
A static tactility-based gait recognition method used in entrance guards and checking-in systems was presented. The static plantar pressure image of a man＇s bare feet collected through a pressure board was zoned, denoised and screened to eliminate the influence on the feature extraction. Secondly, the coordinates and pressure value of peak pressure point and pressure center points were extracted according to the different part of the plantar. Then the Laplace spectrum feature was extracted and the plantar pressure ratio and shape characteristics were also composited in order to improve the algorithm＇s robustness. Finally, the classifier was trained by the classification method of one-to-one support vector machine （SVM） on the trained samples, and 50 person＇s plantar pressure images were selected as the test samples by this gait recognition experiment. Results show that average recognition of the algorithm can achieve 96 % with low interference.
同期刊论文项目
同项目期刊论文
Remote sensing inversion of leaf area index based on support vector machine regression in winter whe
期刊信息
位置：