| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.94 MB | Adobe PDF |
Advisor(s)
Abstract(s)
A dissertação apresentada foi desenvolvida com o intuito de aliar tecnologia de processamento de imagem com mecanismos de decisão e cinemática. Desta forma, utilizando um ambiente de interação com o utilizador pode-se dar resposta ao mesmo de forma autónoma.
O ambiente escolhido é o tradicional jogo do galo, este é composto por um tabuleiro com 9 casas e destinado a 2 jogadores. O jogo começa com a escolha da marca a utilizar por cada jogador, X ou O em seguida, alternadamente, cada jogador deve preencher uma casa do tabuleiro com a sua marca sendo a vitória atribuída ao jogador que conseguir colocar três peças em linha com a sua marca, esta pode ser vertical, horizontal ou diagonal.
Para a interação do computador com o jogador vai ser utilizado um manipulador para traduzir fisicamente a resposta calculada pelo computador à jogada do utilizador. Neste âmbito entram em jogo duas disciplinas, muito abrangentes e cada vez mais importantes no desenvolvimento tecnológico mundial, robótica e processamento de imagem.
No campo do processamento de imagem tem-se como tarefas: identificação do estado do ambiente de interação, processamento da jogada do utilizador e recolha da posição física das peças colocadas em jogo. Aliado a este processo existe um algoritmo que com base no estado do jogo irá calcular a melhor resposta a efetuar pelo manipulador.
De forma a conseguir uma recolha correta do ambiente de jogo, foram utilizadas rotinas de processamento de imagem com métodos para reduzir o ruído que possa ser introduzido por variáveis externas ou alterações no ambiente e corretamente identificar a peça introduzida em jogo.
Após a correta análise do estado atual do tabuleiro de jogo e com base na jogada calculada para o computador é determinado o movimento a executar pelo manipulador com recurso a cinemática. Com a informação necessária para realizar o movimento da peça a mesma é traduzida com recurso a um sistema dedicado ao seu tratamento e compatibilização com o manipulador.
Em seguida os movimentos de cada junta previamente determinados são
convertidos para grandezas analógicas através dum processador Arduíno e enviados para o manipulador.
The dissertation presented was developed with the aim of combining computational vision technology with decision-making mechanisms and kinematics. In this way, using an interaction environment with the user, we will be able to respond autonomously. The environment chosen is the traditional tic-tac-toe game, it consists of a board with 9 fields and intended for 2 players. The game begins with the choice of the mark that will used by each player, X or O then alternately, each player must fill a field on the board with their mark, with the victory being attributed to the player who manages to place three pieces in line with their mark, this line can be vertical, horizontal or diagonal. For the interaction between the computer and the player, a manipulator will be used to physically translate the response calculated by the computer. In this context, two disciplines come into play both being very comprehensive and increasingly important in the current worlds’ technological development, robotics and image processing by means of computational algorithms. In the field of computational vision, we have the following tasks: identifying the state of the interaction environment, processing the user's move and collecting the physical position of the pieces placed in play. Allied to this process, there is an algorithm that based on the state of the game will calculate the best answer for the computer. To obtain a correct collection of the game environment, image processing routines were used with methods to filter the noise that can be introduced by external variables or ambient changes and correctly identify the piece introduced in game. After the correct analysis of the current state of the board, based on the calculated move for the computer the movement to be performed by the manipulator is determined using kinematics, after this the information needed to move the piece is translated using a system dedicated to its treatment and compatibility with the manipulator. Then the previously determined movements of each joint are converted to analog values through an Arduino processor and sent to the manipulator.
The dissertation presented was developed with the aim of combining computational vision technology with decision-making mechanisms and kinematics. In this way, using an interaction environment with the user, we will be able to respond autonomously. The environment chosen is the traditional tic-tac-toe game, it consists of a board with 9 fields and intended for 2 players. The game begins with the choice of the mark that will used by each player, X or O then alternately, each player must fill a field on the board with their mark, with the victory being attributed to the player who manages to place three pieces in line with their mark, this line can be vertical, horizontal or diagonal. For the interaction between the computer and the player, a manipulator will be used to physically translate the response calculated by the computer. In this context, two disciplines come into play both being very comprehensive and increasingly important in the current worlds’ technological development, robotics and image processing by means of computational algorithms. In the field of computational vision, we have the following tasks: identifying the state of the interaction environment, processing the user's move and collecting the physical position of the pieces placed in play. Allied to this process, there is an algorithm that based on the state of the game will calculate the best answer for the computer. To obtain a correct collection of the game environment, image processing routines were used with methods to filter the noise that can be introduced by external variables or ambient changes and correctly identify the piece introduced in game. After the correct analysis of the current state of the board, based on the calculated move for the computer the movement to be performed by the manipulator is determined using kinematics, after this the information needed to move the piece is translated using a system dedicated to its treatment and compatibility with the manipulator. Then the previously determined movements of each joint are converted to analog values through an Arduino processor and sent to the manipulator.
Description
Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia Eletrotécnica – Ramo de Automação e Eletrónica Industrial
Keywords
Interação Humano-Computador Processamento de Imagem Arduino Cinemática Human-Machine Interaction Computational Vision Kinematics
Citation
ORVALHO, Rúben Alexandre de Gonçalves – Jogo do Galo com o Robot. Lisboa: Instituto Superior de Engenharia de Lisboa. 2023. Dissertação de Mestrado.
Publisher
Instituto Superior de Engenharia de Lisboa