Motion Recognition And Software Realization Of Racket Motion

With the improvement of people's living standards,people pay more and more attention to sports and fitness.More and more people like to exercise through exercise,especially the racquet sports(badminton,table tennis,tennis,etc.)are deeply popular among the people.Love.At the same time,with the development of smart phones and sensor networks,wearable smart sports equipment such as sports bracelets,smart watches,and smart sports shoes have also developed rapidly.Most of these smart wearable devices count the relevant data during the movement of the athletes.These statistics can be used to visually show the movement characteristics of the athletes.In addition,the analysis of these statistics can be used to guide the movement of the athletes.Improve the athletic level of the athletes and make the athletes more interested in sports.However,at present,there are relatively few studies on the recognition of racquet sports swing motion.Therefore,this article takes badminton as an example,using a single inertial sensor to collect 6 kinds of common badminton swing motion data,and research a badminton swing.The action recognition algorithm,and based on this,developed a smart badminton training system software that integrates badminton swing action real-time recognition,sports statistics,motion analysis and sharing functions.The main research contents are as follows:(1)A recognition method for badminton swing motion based on single inertial sensor is proposed.In this paper,we use a data acquisition method that fixes a single inertial sensor at the bottom of a badminton racket.A window segmentation method based on real-time motion data stream combining sliding window and action window is proposed.Then the captured motion data is extracted and used.The algorithm performs dimensionality reduction.Finally,a two-layer classifier-based recognition algorithm is proposed to identify six common swing actions.The first layer recognition algorithm uses the C4.5 decision tree algorithm to identify the grip mode of the athlete.The second layer recognition algorithm is based on randomization.The forest algorithm recognizes the swing action.The final experiment shows that the accuracy of the recognition of the common six types of swing motions is 95.36%.(2)Designed and implemented an intelligent badminton training system software.Based on the above-mentioned badminton swing motion recognition algorithm,this paper develops an intelligent badminton training system based on Android mobile operating system.The developed software consists of a BLE(Bluetooth Low Energy)data acquisition module,a registration login module,a motion statistics and analysis module,and a social module.It can realize the real-time recognition of badminton swing action,the statistics and analysis of sports,the sharing of sports status,etc.,which can be used for auxiliary training.