摘要
量化运动能耗是科学运动的基础。目前日常运动能耗的检测设备存在诸多限制条件。针对这一问题,采用近些年飞速发展的图像识别技术,设计了一套基于移动手机端的人体运动能耗检测系统,系统包含4种运动健身动作,通过手机前置摄像头拍摄在语音和视频指导下人体完成相应动作的视频,利用图像分析技术实时计算出人体运动能耗值,实现对人体运动时的能耗检测,实验结果与被称为"金标准"的间接测热法测得值的相关系数均达到了0.6以上(P<0.01),另外提出了4种动作的能耗预测方程,采用Bland-Altman分析法分析验证组的实验结果与标准值得一致性程度,结果表明两种方法的一致性较高,可作为传统方法的替代。
Quantifying sports energy consumption is the foundation of scientific sports. At present, there are many limitations in the detection equipment of daily exercise energy consumption. To address this problem, this paper adopts the image recognition technology that has been developing rapidly in recent years, and designs a set of human movement energy consumption monitoring system based on the mobile phone terminal, the system contains four kinds of movement and fitness actions, through the mobile phone front camera to shoot the video of the human body completing the corresponding action under the voice and video guidance, using the image analysis technology to calculate the human movement energy consumption value in real time, to realize the human movement energy consumption monitoring. The correlation coefficients between the experimental results and the measured value of indirect calorimetry, which is called the "gold standard", reached more than 0.6(P<0.01), and the energy consumption prediction equations for four actions were proposed. The methods are more consistent and can be used as an alternative to traditional methods.
作者
陈超
孙少明
王威
陈竟成
张海涛
Chen Chao;Sun Shaoming;Wang Wei;Chen Jingcheng;Zhang Haitao(Hefei Institutes of Physical Science,Chinese Academy of Sciences,Hefei 230031,China;University of Science and Technology of China,Hefei 230026,China;CAS(Hefei)Institute of Technology Innovation,Hefei 230088,China)
出处
《电子测量技术》
北大核心
2021年第2期115-120,共6页
Electronic Measurement Technology
基金
中国科学技术大学智慧城市研究院(芜湖)科技成果转化项目(2019ZX01)
国家重点研发计划(2018YFC2001304)项目资助。
关键词
移动端
能耗测量
动作检测
mobile
energy consumption measurement
motion detection
作者简介
陈超,在读硕士研究生,主要研究方向为智能检测技术。E-mail:ahcc0920@163.com;通信作者:孙少明,研究员,博士,主要研究方向为运动与健康。E-mail:smsun@iim.ac.cn。