Publication: Desarrollo de un sistema de juego al Tres en Raya para el robot NAO H25
Loading...
Identifiers
Publication date
2013-06-25
Defense date
2013-07-04
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
En este trabajo se presenta el desarrollo de un módulo para el robot NAO que le permitirá jugar
al “Tres en Raya” con un humano de forma autónoma.
El desarrollo del juego incluye una gran variedad de técnicas de inteligencia artificial como
el aprendizaje por demostración, la visión artificial y la búsqueda heurística que le permitirán
alcanzar al robot ese grado de autonomía. La finalidad del trabajo es conseguir que el robot
NAO aumente el componente social e interactúe cada vez mejor con los humanos mediante
juegos sencillos que todos conocemos, pues este trabajo sirve de base para desarrollar otros
módulos con otros juegos diferentes.
En la realización del proyecto se ha utilizado el simulador Choregraphe y el API de NaoQi, desarrollados
por Aldebaran Robotics para trabajar expresamente con el robot NAO y las librerías
de programación OpenCV, para el procesamiento de imágenes, y AIMA, para el algoritmo del
juego. El código se ha desarrollado en Python. El proyecto se ha probado sobre un entorno real,
con un jugador humano y un robot NAO.
This thesis presents the design and development of a new module for NAO, the humanoid robot, which allows it to play Tic Tac Toe against a human player in an autonomous way. The development includes a wide variety of artificial intelligence based techniques such as Learning From Demonstration, Computer Vision and Heuristic Search in order to reach that autonomous behavior on the robot. The aim of this work is to get the robot NAO increase its social component to improve the interaction with humans through simple well-known games. This work serves as the basis to develop other modules with different games. The code of the Tic Tac Toe module has been developed in Python with the help of Aldebaran’s Choregraphe simulator and NaoQi API to work with the robot, OpenCV for image processing and AIMA library for the algorithm of the game. The module has been tested in a real environment where a Tic Tac Toe game was played by a human and a robot NAO.
This thesis presents the design and development of a new module for NAO, the humanoid robot, which allows it to play Tic Tac Toe against a human player in an autonomous way. The development includes a wide variety of artificial intelligence based techniques such as Learning From Demonstration, Computer Vision and Heuristic Search in order to reach that autonomous behavior on the robot. The aim of this work is to get the robot NAO increase its social component to improve the interaction with humans through simple well-known games. This work serves as the basis to develop other modules with different games. The code of the Tic Tac Toe module has been developed in Python with the help of Aldebaran’s Choregraphe simulator and NaoQi API to work with the robot, OpenCV for image processing and AIMA library for the algorithm of the game. The module has been tested in a real environment where a Tic Tac Toe game was played by a human and a robot NAO.
Description
Keywords
Juegos de ordenador, Robots, Inteligencia artificial, Aprendizaje