摘要
针对太极拳训练方式单一和效率低、单个Kinect捕捉范围有限且存在数据遮挡等问题,文中提出多Kinect下的太极拳辅助训练系统。该系统利用Kinect获取到人体彩色视频和深度数据,分别获得两个Kinect下用户的运动数据。由于两个Kinect下获取的人体骨骼关节点都是相对于当前坐标系的,因此利用最小二乘法进行坐标标定,再利用卡尔曼滤波算法进行数据融合。就得到融合后的数据。利用动态时间规划算法匹配用户的动作序列与标准序列,对比其角度、坐标点、时间等特征的不同,结合相关规则进行评分。该研究内容对传统体育教学和科学训练有重要借鉴意义。
Aiming at the problems of single Taijiquan training method,low efficiency,limited capture range of a single Kinect and data occlusion,a Taijiquan auxiliary training system under multiple Kinects is proposed.The system uses Kinect to obtain human color video and depth data,and obtains the user's motion data under two Kinects.Because the joint points of the human bones acquired under the two Kinects are all in the current coordinate system.Therefore,the least square method is used for coordinate calibration,and then the Kalman filter algorithm is used for data fusion.Get the data after fusion.The dynamic time planning algorithm is used to match the user's action sequence with the standard sequence,and the differences in the angle,coordinate point,time and other characteristics are compared,and the relevant rules are combined to score.The content of this research is of great significance to traditional physical education and scientific training.
作者
闵亚荣
牟莉
MIN Yarong;MOU Li(College of Computer Science,Xi’an Polytechnic University,Xi’an 710048,China)
出处
《电子设计工程》
2021年第18期115-120,130,共7页
Electronic Design Engineering
关键词
KINECT
辅助训练
坐标标定
数据融合
动作识别
Kinect
auxiliary training
coordinate calibration
data fusion
action recognition
