Initial Approach on Soccer Match’s Scene Classification by Players’ Field Spatial Distribution

 

Сохранить в:
Библиографические подробности
Авторы: Núñez-Meoño, Lennon, Villalta, Marco, Siles-Canales, Francisco
Формат: artículo original
Статус:Versión publicada
Дата публикации:2020
Описание:Soccer player tracking is moving towards the use of high-cost algorithms; however, these are implemented to the whole video recording of interest with no flexibility on the type of scene being evaluated. The current work proposes to classify the scenes according to the players’ distribution over the field, designing a metric for players’ Dispersion Level that would allow implementing a more flexible tracking method that selects a light algorithm for well-behaved scenes and a heavy and robust algorithm for bad-behaved scenes. This work pushes towards considering different existent metrics, as Voronoi diagrams for surface area analysis, in future works to come.
Страна:Portal de Revistas TEC
Институт:Instituto Tecnológico de Costa Rica
Repositorio:Portal de Revistas TEC
Язык:Inglés
OAI Identifier:oai:ojs.pkp.sfu.ca:article/5077
Online-ссылка:https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/5077
Ключевое слово:Soccer player
object tracking
scene classification
occlusion
stretch index
player dispersion
surface area
Jugador de fútbol
rastreo de objetos
clasificación de escenas
oclusión
índice de estrechés
dispersión de los jugadores
área superficial