基于三轴加速度传感器的人体动作识别研究

摘要

提出一种基于三轴加速度传感器的人体动作识别方法，通过对采集到的数据进行预处理，提取多种统计特征包括标准差、阈值、偏度、峰度等进行分类识别。能够有效地识别站立、行走、跑、跳这四种人体基本日常行为动作。%In this paper,a method for activity recognition based on 3D acceleration sensor is introduced.After sensor data are col ected,preprocessing,many statistical features are extracted such as standard deviation,threshold,skewness and kurto-sis for classification.The system is trained to recognize four human daily activities:staying,walking,running and jumping.