基于深度学习的羽毛球正手高远球动作标准智能评估方法

打开文本图片集

文章编号：1003-6180（2026）02-0050-07

[中图分类号]TP391

[文献标志码]A

An Intelligent Evaluation Method for the Standard of Forehand High and Far Shots in Badminton based on Deep Learning

TONG Chuan

（College of police command and tactics， Anhui Police College， Hefei ， China）

Abstract： Propose a deep learning based intelligent evaluation method for high and far ball movements of badminton forehand （DL-BEFBFS）. Input image data samples into an improved convolutional neural network， extract spatiotemporal features of badminton movements， and obtain intelligent recognition results of badminton forehand movements; Using the spatiotemporal characteristics of actions to calculate intelligent evaluation indicators such as frame height matching degree and center of gravity stability index; Assign weights to indicators based on their contribution， and use principal component analysis to achieve intelligent evaluation of badminton forehand high and far ball action standards. The experimental results show that the action level evaluation accuracy of the DL-BEFBFS method is 0.97， which is 0.075 higher than traditional evaluation methods and has better evaluation performance.

Keywords： deep learning algorithm b; adminton moves; intelligent assessment

羽毛球正手高远球是后场击球技术中的核心动作，其技术体系涵盖准备姿势、发力机制、击球控制及随动调整四大环节，通过身体协调发力实现球的垂直下落与远距离飞行。（剩余5958字）