A model for Managing the Competitive Activitiesof Top-Level Teams Based on Computer Vision Online

The use of PIRS (Polozov Information Rating System) technology in working with the Russian men’s national futsal team (head coach S.L. Skorovich) in the period from 2010 to 2018 allowed it to move from 5th place in the world ranking to first. The problem of managing a high-level sports team is to selectthe control indicators and their values in order to get the best result of the next match. Usually such parameters are: optimization of the composition, the direction of the game, the placement of their players, the tasks recommended to them. The aim of the study was to structure the activities for the transition from PIRS technology in working with top-level teams to the creation of the PIRSonline system. The description of the PIRSonline system for the development of its technical part required obtaining a number of schemes with their description, which regulated the development of the project. This is the composition of documentation, subsystems, functional diagram, component diagram, function-component matrix. The resulting sequence of actions for the development of PIRSonline includes image acquisition, a block for initializing game parameters, ball detection, determining the 3D coordinates of the ball, detecting players, tracking, classifying players and referees, determining the player’s pose. This made it possible to implement an integrated approach to online management of competitive activities of top-level teams. The use of modern computer vision and machine learning methods has made it possible to create interactive data visualization.