®®®® SIIA Público

Título del libro: Proceedings Of The 2015 International Conference On Artificial Intelligence, Icai 2015 - Worldcomp 2015
Título del capítulo: Determining humanoid soccer player position based on Goal detection

Autores UNAM:
VICTOR MANUEL LOMAS BARRIE; JUAN MARIO PEÑA CABRERA; ADALBERTO JOEL DURAN ORTEGA;
Autores externos:

Idioma:
Inglés
Año de publicación:
2019
Palabras clave:

Computer games; Error detection; Neural networks; Boundary object functions; Fuzzy ARTMAP; Goal detection; Humanoid; Position; RoboCup soccer; Soccer; University teams; Sports


Resumen:

Find the goal in a game from the RoboCup Soccer Humanoid League using computer vision is one of the main tasks in this kind of competitions. Detect the contrary team's goal it is important, if one of the humanoid team members seeks its own goal the robot could share the information with the rest of the members team. In this paper, the Humanoid University team focus in a method which not only detect the contrary or own goal but also find the humanoid relatively position to the goal. The target is extract a goal descriptor vector (Boundary Object Function method) from different points of view (a goal's point-ofview map). With all of those vectors an Artificial Neural Network FuzzyARTMAP is trained. Thus, using the classification FuzzyARTMAP process, each humanoid can aim to the goal and figure it out where is itself located. © 2019 ICAI 2015 - WORLDCOMP 2015. All rights reserved.


Entidades citadas de la UNAM: