Algorithm Research And Software Implementation Of Badminton Swing Movement Recognition

Since the beginning of the 21 st century,people’s material life has gradually enriched.Many people no longer only pay attention to food and clothing.Their health and quality of life are gradually being cared by people.More and more people are keen on physical fitness.It is popular with people because of its interesting and varied features.Thanks to the rapid development of science and technology,smart devices and network technologies such as smart sports bracelets,smart watches,smart insoles,smart helmets and other products have also developed rapidly.After people wear these smart devices,they collect the wearer’s motion data,obtain the wearer’s detailed motion information through the wearer’s motion data,and then feed the information back to the athlete.It enables the athletes to better understand their physical condition and competitive level while exercising,and it can also increase the fun of the exercise.However,there are very few smart devices or statistical systems for badminton.Therefore,this article focuses on this,and proposes an algorithm that can identify the 6 common types of badminton swings,and designs and implements an algorithm that can recognize swings in real time,count sports data,analyze sports data and have social functions.Mobile phone application software.The research content is mainly divided into the following two parts:(1)An action recognition algorithm based on a two-layer classifier is developed,which can recognize the 6 common types of badminton swing actions.In this paper,a single inertial sensor fixed on the bottom of the badminton racket grip is used to collect the swing motion data of 12 players,and use Blue Low Energy(BLE)for data transmission and collection.The collected real-time data is intercepted by the window cutting method combining action window and sliding window,and the action data features after the window interception are extracted,and then Principal Component Analysis(PCA)algorithm and Linear Discriminant Analysis(LDA)algorithm are used to reduce the dimensionality of the feature data to indirectly realize feature selection.Finally,a two-layer classifier based on Support Vector Machine(SVM)and Adaptive Boosting(Ada Boost)is used to identify the 6 common swing movements in badminton.Experiments have proved that the method used in this paper can achieve 96.75%recognition rate of the 6 common badminton swing motion types.(2)Developed an i OS-based badminton auxiliary training application software.The software communicates with data acquisition equipment through BLE to collect real-time swing motion data of athletes,and can identify specific swing motion types in real time.In addition,the application software also includes registration and login,statistical analysis of sports data and social functions.Users can share exercise data and exercise status with other friends who exercise together through the social module,and learn about each other to help athletes improve their competitive level.