Development Of Intelligent Sports Training System Based On Vision Sensing

With the rapid development of Internet technology and somatosensory technology,people's growing needs of physical health have promoted the transformation of the sports industry.The latest science and technology have entered the sports industry and continue to impact traditional sports equipment.However,at present,the technical programs of traditional sports equipment in colleges and universities are unreasonable,and the management model of sports training is outdated,it is impossible to scientifically guide the student movement,and the current situation of overall sports performance is poor.This thesis proposes an intelligent sports training system based on visual sensing.Aiming at the problem of research on the field of motion recognition is staying in recognition of single action,this paper is devoted to the research of continuous motion recognition,uses hidden Markov model to modeling the sequence with time continuity in the process of motion change,and an identification algorithm for efficient detection of motion is proposed.Aiming at the problem of sports equipments in colleges and universities are simple and have single function,a terminal with multi-functional recognition of motion based on cloud monitoring is proposed.The cloud server communicates and transmits data with the recognition of motion through HTTP,manages the monitoring terminal to achieve terminal positioning,sports data collection,etc.The sports terminal have realized multi-functional sports training such as sit-ups,pull-ups,squats,and standing long jumps.Aiming at the problems that the traditional sensor and other wearable sports equipment are difficult to unify management,such as cumbersome operation,easy to damage and need manual counting and so on,a non-wearing unsupervised sports training mode based on visual sensing is proposed.The motion terminal collects the video data stream of the user motion,extracts the feature and establishes the HMM algorithm for recognition of motion to realize the recognition of the motion action,the automatic counting and the intelligent function of scoring.Aiming at the lack of scientific guidance of sports training program for college students,which resulting in the overall sports performance is poor,the customization of scientific sports training program based on BMI index is proposed.The international general health BMI index is used to analyze the sports data uploaded by a large number of students in real time,so as to realize a personalized scientific sports training plan according to different student physiques.The system has been developed and tested with stable and reliable performance.It has been put into use in colleges and universities.After test of practical application,the intelligent training system for motion based on visual sensing proposed in this paper has both practicality and interest,and has been recognized by the market.