The paper focuses on the relevance of three-axis accelerometers and the badminton swing movement coordination and stability. Some of the studies used recording devices or optical instrument to analyze the player's actions, this method is limited by the venue, as well as photography dead, privacy issues, cost, complexity is high. In this paper, the method can effectively solve these problems. The methods uses less computing complexity, achieve effective analysis badminton swing action. After obtaining the three-axis acceleration value, use low pass filtering process for the noise. The characteristic values of different parts of the swing use neural networks to classify the action into error and correct operation provides different feedback for user. We analyzed the data of badminton clear. In the future, we will analysis other strokes in badminton.
展开▼
