The concept of shot put, the structure of the shot put sector, the study of the results of the shot putters of Ukraine and the world arena, the description of the shooting technique "from the jump", the use of the concept of computer vision in sports are considered. The models and methods for calculating the performance of a shot putter athlete, models for calculating the shot put "from the spot" and "from the jump", the main tasks for the apparatus of neural networks and the use of the concept of computer vision to create a video analysis system are analyzed. The main task and purpose of the work is set. The formalization and algorithm of the neural network model for evaluating the pushing phases is presented. The created information model of the designed system is described in the language of visual modeling UML - diagrams of use cases, classes, cooperation, sequence, states and components are given. The possibilities of the system for studying the main indicators of a shot putter athlete, the possibility of using a video system to improve the technique of pushing are described. An example of the operation of this system is given and an analysis of the calculation results is carried out.
Read full abstract