Application of Direction Vector Model of Multibody System Dynamics in Sports Posture Positioning.

Conventional sports posture positioning method, using convolution neural network, only uses Fast RCNN algorithm to locate sports posture, which leads to low accuracy of sports posture positioning. This paper proposes the application of multibody system dynamic direction vector model in sports posture positioning. In this paper, the direction vector model of multibody system dynamics is used to design the constraint of direction vector property and hinge between objects, calculate the activity of objects, and establish the motion posture model; The four coordinate systems of image, pixel, camera and world are established, and the spatial relationship of the four coordinate systems is calculated to complete the spatial transformation of each coordinate system; The gray transformation is carried out for the image to obtain an image. Gray human body image is used to obtain the target image and extract the contour of human body region; Similar to Fast RCNN algorithm and multi-scale context information are used to form a multi scale convolution neural network to locate the sports posture. The experimental results show that positioning sports posture has better positioning effect, higher positioning accuracy and faster positioning efficiency. [ABSTRACT FROM AUTHOR]