發(fā)布時間：2018-04-15 10:18

  本文選題：體育訓練 + 運動視頻��； 參考：《現(xiàn)代電子技術》2017年11期

【摘要】：當前的體育訓練過程中,訓練員使用難度較小的視頻重播與解析管理方式為運動員講解動作要領,不夠直觀和科學,不能滿足訓練員對運動效果評估的需求。針對該問題,研究了體育訓練過程中的運動視頻分析與識別過程,采用基于粒子濾波預測的自適應閾值運動目標分離算法實現(xiàn)運動目標的自適應分離。通過粒子濾波技術跟蹤運動員的運動,塑造運動模型,并依據(jù)運動模型預測后續(xù)運動視頻幀內(nèi)不同重要關節(jié)點的位置,完成后續(xù)運動視頻幀的跟蹤。采用條件隨機場方法實現(xiàn)體育訓練視頻中的動作識別。實驗結果說明該方法具有較高的動作識別率和較低的誤分離率。
[Abstract]:In the current sports training process, the trainers use the less difficult video replay and analysis management to explain the main points of action for the athletes, which is not intuitive and scientific enough to meet the needs of the trainers for the evaluation of sports effects.In order to solve this problem, the process of motion video analysis and recognition in the process of sports training is studied, and the adaptive threshold motion target separation algorithm based on particle filter prediction is used to realize the adaptive separation of moving objects.The particle filter technique is used to track the movement of athletes, to shape the motion model, and to predict the position of different key points in the frame of the subsequent motion video according to the motion model, so as to complete the tracking of the following motion video frame.The method of conditional random field is used to realize motion recognition in sports training video.The experimental results show that the method has higher recognition rate and lower error separation rate.
【作者單位】： 黃河科技學院;
【基金】：河南省科技廳科技攻關項目(132102310462)
【分類號】：TN713;TP391.41
，

本文編號：1753696