本文旨在完善羽毛球运动员反手高远球的技术动作、纠正羽毛球技术动作的易错点，提高教学质量，研发了姿态识别系统，采用了一种基于OpenPose的人体姿态估计方法，通过估计身体各部位的坐标和角度，将运动员的姿态信息转化为可处理的数字特征，进而提取出反手高远球动作的关键帧，基于系统采集出的信息形成运动学数据，对数据进行分析与反馈，以提高技术动作的量化评价。为了验证本研究方法的有效性和可行性，我们对大量的羽毛球反手高远球动作数据进行了收集和标注，并进行了实验验证。实验结果表明，该方法可以实现对羽毛球反手高远球动作的准确识别和评价，同时也具有较高的实用性和可扩展性，实现了对羽毛球反手高远球动作的自动识别和评价。该方法可以有效提高羽毛球运动员的训练效率和技能水平，可用于羽毛球教学系统中的教学评价，对羽毛球运动的发展具有重要的实用价值。

This article aims to improve the technical movements of badminton players in backhand high and long shots,correct the prone points of technical movements in badminton,and improve teaching quality.A posture recognition system has been developed,and an OpenPose based human posture estimation method has been adopted.By estimating the seating positions and angles of various parts of the body,the athlete's posture information is transformed into processable digital features,and the key frames of backhand high and long shot movements are extracted,The kinematics data is formed based on the information collected by the system,and the data is analyzed and fed back to improve the quantitative evaluation of technical actions.In order to verify the effectiveness and feasibility of this research method,we collected and annotated a large amount of data on backhand and long ball movements in badminton,and conducted experimental verification.The experimental results show that this method can achieve accurate recognition and evaluation of backhand high and far ball movements in badminton,and also has high practicality and scalability,achieving automatic recognition and evaluation of backhand high and far ball movements in badminton.This method can effectively improve the training efficiency and skill level of badminton athletes,and can be used for teaching evaluation in the badminton teaching system.It has important practical value for the development of badminton sports.