Research On Archery Athlete Posture Parameters Acquisition Method In Video Stream

In recent years the global public take attention to the sports competition. Improving thelevel of sports competition become the ultimate goal of all countries of exercise training.Conform to the requirements of the epoch, Sports training methods break the traditionalpattern which training by the way of words and deeds from the coach and combine withscience, which formed a comprehensive training that integrated into the elements of scienceand technology. The analysis of science and technology about athletes training is ahigh-profile auxiliary training method between coaches and athletes..Among the many sports analysis methods，to obtain human motion parameters byvideo images is the most effective method available. The video stream contains a largenumber of human motion information whose performance of the Real-time analysis hasbeen further improved, during the training, the analysis of results and feedback data can beprovided timely. Coaches can clearly point out the shortcomings and deficiencies of presenttechnical level in the presence of the athletes and give the corresponding improvementsthrough combining domain knowledge of specific sport and quantitative information ofvideo data. Archery athletes can also do targeted quantization training and digitalize of thetraining target so that they can improve the level and efficiency of training rapidly.Through analysis of the existing algorithms and combined with the need of archeryathlete training, this paper proposes an algorithm based on adaptive threshold of Otsu forobject detecting. It added a constraint between Otsu. The results can be guaranteed to remainintact prospect information. To solve mufti-target tracking problems, this paper proposes amufti-target tracking algorithm based on a combination of corner feature. It extracts stableand symmetrical feature points of moving object using the improved Harris operator, andcompletes the tracking of video moving mufti-target by feature matching and matchingoptimization. Tracking experiments show that the algorithm can complete stable matchingunder the changes of angle view, rotation, affine transformation, illumination and other circumstances, and can achieve stable tracking under a small partial shelter state. Accordingathlete upper centroid position changing, this paper proposes an algorithm based on centroidweighted Kalman filter for object tracking. The algorithm firstly uses backgroundsubtraction method to lock dynamic target tracking area, and then uses the Kalman filter topredict the target’s position at the beginning of the target tracking, and then optimizes thepredictive state value adopting centroid weighted method, finally updates the observationdata according to the corrected state value. Simulation results show that the algorithm candetect effectively moving objects and at the same time it can quickly and accurately trackmoving objects with good robustness.Finally, acquisition and analysis parameters for motionstance of archery in this paper provided reliable parameter analysis for the coaches andathletes.