Machine Vision Based Positioning Technique for Sports 3D Action Oriented Teaching

In this talk, a sport 3-dimensional (3D) teaching action positioning method is discussed. It utilizes machine vision technologies to solve the problems of low positioning accuracy and long positioning time that are used in a traditional sport 3D teaching action positioning method. Step-by-step working of the proposed method is as follows: (a) world coordinate system of the images are determined by machine vision method; (b) the MRF algorithm is used to extract the features of the 3D teaching action image; and (c) the spatial correlation of the 3D teaching action data is determined. Smoothing filtering technology is used to suppressed and eliminate noise in the 3D teaching action image. 3D teaching action image is then reconstructed using the convolutional neural network (CNN). CNN is used to determine the entropy of sports 3D teaching action. This talk describes three key elements are as follows - (a) how 3D sports teaching action image is obtained through the camera and then located the action by using the artificial intelligence algorithm; (b) how MRF algorithm is used to extract the features of 3D teaching action image of physical education and to determine the spatial correlation of 3D teaching action data of physical education; and (c) how a CNN is used to determine the entropy value of sports 3D teaching action image and to realize the positioning of sports 3D teaching action. Extensive computer simulations are conducted to determine the effectiveness of the proposed system. Result reveals that the proposed system achieved 95% accuracy with 1.9s of the positioning time. Comparative results and discussions against the state-of-the art methods are highlighted. Finally, this talk will focus on application area and future directions.