基于信号处理和机器学习的人类活动识别研究  

Research on Human Activity Recognition Based on Signal Processing and Machine Learning

作　　者：殷建艳 YIN Jianyan(Tangshan Preschool Teachers College,Tangshan 063000,China)

机构地区：[1]唐山幼儿师范高等专科学校,河北唐山063000

出　　处：《工业技术与职业教育》2025年第2期1-7,共7页Industrial Technology and Vocational Education

摘　　要：人类活动检测是行为识别领域的重要任务,其中跌倒检测因其安全性和紧急干预需求而备受关注。结合信号处理技术与机器学习模型,对人类活动检测任务进行了系统研究,并重点分析了跌倒检测的性能表现。实验结果表明,各模型均表现良好,并在不同任务中展现出各自的优势。为跌倒检测系统的开发提供了重要参考,同时为活动检测模型的优化奠定了基础。Human activity detection is an important task in behaviour recognition,with fall detection being particularly noteworthy due to its safety concerns and emergency intervention requirements.This study combines signal processing techniques with machine learning models to systematically investigate human activity detection tasks,with a particular focus on analysing the performance of fall detection.The experimental results show that all models perform well and exhibit their respective advantages in different tasks.This research provides important references for the development of fall detection systems and lays the foundation for the optimisation of activity detection models.

关 键 词：信号预处理 机器学习 人类活动识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]