计算机应用 ›› 2019, Vol. 39 ›› Issue (5): 1448-1452.DOI: 10.11772/j.issn.1001-9081.2018102137

• 虚拟现实与多媒体计算 • 上一篇    下一篇

基于人体骨架的非标准深蹲姿势检测方法

喻露1, 胡剑锋1,2, 姚磊岳2   

  1. 1. 南昌大学 信息工程学院, 南昌 330031;
    2. 江西科技学院 协同创新中心, 南昌 330098
  • 收稿日期:2018-10-24 修回日期:2019-01-02 发布日期:2019-05-14 出版日期:2019-05-10
  • 通讯作者: 喻露
  • 作者简介:喻露(1994-),男,江西宜春人,硕士研究生,主要研究方向:计算机视觉、行为识别、目标跟踪;胡剑锋(1976-),男,江西景德镇人,教授,博士,主要研究方向:神经网络、脑电波、机器学习;姚磊岳(1982-),男,浙江舟山人,教授,博士,主要研究方向:计算机视觉、信息处理。
  • 基金资助:
    国家自然科学基金资助项目(61762045);江西省科技厅项目(20171BAB202031);江西省科技厅科技攻关项目(20171BBE50060);江西省博士后援助项目(2017KY33);江西省教育厅项目(GJJ161143,GJJ151146);江西省科技厅科技计划专项重点研发项目(20181BBE50018);南昌市科技局科技规划项目(2016-ZCJHCXY-013)。

Detection method of non-standard deep squat posture based on human skeleton

YU Lu1, HU Jianfeng1,2, YAO Leiyue2   

  1. 1. Information Engineering School, Nanchang University, Nanchang Jiangxi 330031, China;
    2. Center of Collaboration and Innovation, Jiangxi University of Technology, Nanchang Jiangxi 330098, China
  • Received:2018-10-24 Revised:2019-01-02 Online:2019-05-14 Published:2019-05-10
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61762045), the Project of Science and Technology Department of Jiangxi Province (20171BAB202031), the Science and Technology Research Project of Jiangxi Science and Technology Department (20171BBE50060), the Postdoctoral Assistance Project of Jiangxi Province (2017KY33), the Project of Department of Education of Jiangxi Province (GJJ151146, GJJ161143), the Science and Technology Plan Special Key Research and Development Project of Jiangxi Science and Technology Department (20181BBE50018), the Science and Technology Planning Project of Nanchang Science and Technology Bureau (2016-ZCJHCXY-013).

摘要: 针对健身者在健身过程中因缺乏监督指导而导致姿势不正确甚至危及健康的问题,提出了一种深蹲姿势实时检测的新方法。通过Kinect摄像头提取人体关节三维信息,对健身中最常见的深蹲行为进行抽象与建模,解决了计算机视觉技术对于细微动作变化难以检测的问题。首先,通过Kinect摄像头捕获深度图像,实时获取人体关节点的三维坐标;然后,将深蹲姿势抽象为躯干角度、髋部角度、膝部角度和踝部角度,并进行数字化建模,逐帧记录下角度变化;最后,在深蹲完成后,采用阈值比较的方法,计算一定时间段内非标准帧比率。如计算比率大于所给定阈值,则判定此次深蹲为不标准;如低于阈值则为标准深蹲姿势。通过对六种不同类型的深蹲姿势进行实验,结果表明,该方法可检测出不同类型的非标准深蹲姿势,并且在六种不同类型的深蹲姿势中平均识别率在90%以上,能够对健身者起到提醒指导的作用。

关键词: 深蹲检测, 姿势检测, Kinect, 深度图像, 骨架信息

Abstract: Concerning the problem that the posture is not correct and even endangers the health of body builder caused by the lack of supervision and guidance in the process of bodybuilding, a new method of real-time detection of deep squat posture was proposed. The most common deep squat behavior in bodybuilding was abstracted and modeled by three-dimensional information of human joints extracted through Kinect camera, solving the problem that computer vision technology is difficult to detect small movements. Firstly, Kinect camera was used to capture the depth images to obtain three-dimensional coordinates of human body joints in real time. Then, the deep squat posture was abstracted as torso angle, hip angle, knee angle and ankle angle, and the digital modeling was carried out to record the angle changes frame by frame. Finally, after completing the deep squat, a threshold comparison method was used to calculate the non-standard frame ratio in a certain period of time. If the calculated ratio was greater than the given threshold, the deep squat was judged as non-standard, otherwise judged as standard. The experiment results of six different types of deep squat show that the proposed method can detect different types of non-standard deep squat, and the average recognition rate is more than 90% of the six different types of deep squat, which can play a role in reminding and guiding bodybuilders.

Key words: deep squat detection, posture detection, Kinect, depth image, skeleton information

中图分类号: