Desenvolvimento de um sistema robótico com movimentos comandados por gestos identificados por visão computacional
Carregando...
Arquivos
Tipo
TCC
Data de publicação
2019-06
Periódico
Citações (Scopus)
Autores
Lozaldo, Bruno Benini
Kolbe, Cauli
Souza, Gabriel Gomes de
Kolbe, Cauli
Souza, Gabriel Gomes de
Orientador
Lima, Bruno Luis Soares de
Título da Revista
ISSN da Revista
Título de Volume
Membros da banca
Programa
Resumo
Atualmente, com o uso crescente de robôs, e pelo elevado nível educacional necessário para programar e interagir com essas máquinas, se faz necessário o desenvolvimento de plataformas que tornem a interação homem-máquina mais simples e precisa. No presente trabalho, desenvolveu-se um sistema de controle para manipuladores robóticos baseado em visão computacional e processamento de imagens. Foi desenvolvido um algoritmo em Phyton baseado na biblioteca de código aberto OpenCV, que identifica comandos gestuais capturados por uma webcam e, a partir de uma interface com o usuário, os transforma em movimentos específicos a serem realizados por um protótipo de manipulador que foi projetado e construído pelos autores. O protótipo, construído em MDF, possui alcance de 180 mm e 3 graus de liberdade, e seus movimentos são realizados a partir do acionamento de 3 motores de passo. O acionamento dos motores se dá a partir de um firmware que foi desenvolvido para o microcontrolador Arduino, o qual recebe o comando gestual que foi detectado. Foi possível estabelecer condições críticas de funcionamento do sistema desenvolvido com relação a luminosidade ambiente, distância do operador em relação a webcam e a precisão mecânica da resposta motora do manipulador. A partir das tecnologias aplicadas neste trabalho poderá ser possível o desenvolvimento de sistemas de controle por gestos para outros mecanismos robóticos.
Nowadays, with the increasing use of robots, and the high level of education required to program and interact with these machines, it becomes necessary to develop platforms that make the human-machine interaction simpler and more precise. In the present work, a control system was developed for robotic manipulators based on computer vision and image processing. An algorithm was developed in Phyton based on the OpenCV open source library, which identifies gesture commands captured by a webcam and, from a user interface, transforms them into specific movements to be performed by a prototype manipulator that was designed and built by the authors. The prototype, built in MDF, has a range of 180 mm and 3 degrees of freedom, and its movements are made from the activation of 3 step motors. The activation of the motors takes place from a firmware that was developed for the Arduino microcontroller, which receives the gestural command that was detected. It was possible to establish critical operating conditions of the developed system in relation to the ambient brightness, distance from the operator to the webcam and the mechanical accuracy of the manipulator's motor response. From the technologies applied in this work it may be possible to develop control systems by gestures for other robotic mechanisms.
Nowadays, with the increasing use of robots, and the high level of education required to program and interact with these machines, it becomes necessary to develop platforms that make the human-machine interaction simpler and more precise. In the present work, a control system was developed for robotic manipulators based on computer vision and image processing. An algorithm was developed in Phyton based on the OpenCV open source library, which identifies gesture commands captured by a webcam and, from a user interface, transforms them into specific movements to be performed by a prototype manipulator that was designed and built by the authors. The prototype, built in MDF, has a range of 180 mm and 3 degrees of freedom, and its movements are made from the activation of 3 step motors. The activation of the motors takes place from a firmware that was developed for the Arduino microcontroller, which receives the gestural command that was detected. It was possible to establish critical operating conditions of the developed system in relation to the ambient brightness, distance from the operator to the webcam and the mechanical accuracy of the manipulator's motor response. From the technologies applied in this work it may be possible to develop control systems by gestures for other robotic mechanisms.
Descrição
Palavras-chave
projeto de manipulador robótico , construção de manipulador robótico , controle gestual , visão computacional com openCV , hotel , costs , processes , finance