Posture recognition method of basketball shooting using image retrieval technology
In the process of basketball shooting,there will be illegal action posture of the batter.There are fac-tors such as the short time of the occurrence of the batter's action posture and the uncertainty of the position of the action posture.There is great difficulty in the recognition.In order to solve the above problems,a basketball shooter's gesture recognition method based on image retrieval technology is proposed.Analyze the characteris-tics of the pixel region formed by the batter's movement posture,use image retrieval technology to find a break-through in the effective pixel feature region,preprocess the image of the edge contour feature of the shooting posture image,and determine the effective region of the image.Then the image features of shooting posture are retrieved,and the pixel features of the beater's posture region are extracted;Finally,the position image recogni-tion sample training is used to realize the recognition of the beater's movement posture.Experimental tests have shown that the proposed method for recognizing the posture of basketball shooters has strong capabilities in improving recognition accuracy,accuracy,and response speed.
basketball shootingposture of the thug's movementsattitude recognitionimage retrieval
NETL
NSTL
万方数据
