体育学员运动姿态判定算法研究

Research on Sports Posture Judgment Algorithm of Sports Students

为了实现对于学员运动姿势的标准化判定,研究基于Openpose工具的学员运动影像判定算法。对Openpose所提供的关节点位置坐标数据进行标准化处理,并将处理结果应用于人体重心矢量角特征判定、肢体夹角特征判定和人体朝向特征判定,进而实现对于学员整体姿态的判定。最后采用所提出的姿态判定算法对10种不同类型的体操动作进行判定,结果表明,该算法能够将误判率控制在10%左右,得到较为理想的判定效果。

In order to judge the posture of the trainees based on the motion vector of the trainees'openpose,and then judge the position of the collected posture of the trainees based on the motion vector of the trainees'openpose.Finally,the posture judgment algorithm proposed in this study is used to judge 10 different types of gymnastics movements.It is found that the algorithm can control the misjudgment rate at about 10%and achieve an ideal judgment effect.