Sistema de visão a laser para mapeamento de superfície de navegação de robôs quadrúpedes

• Autor(a) principal: Orlandi, José Geraldo das Neves
• Data de Publicação: 2008
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Espírito Santo (riUfes)
• Texto Completo: http://repositorio.ufes.br/handle/10/4115
• topic: Robôs móveis; Visão de robô; Navegação de robôs móveis; Processamento de imagens; Desvio de obstáculos; Engenharia Elétrica; 621.3

This work presents, implements and tests a laser vision system to build navigating surface maps and localization of obstacles for quadruped robots. As long as the robot navigates on a pre defined trajectory, the system maps the surface, identifies the obstacles and makes decisions to pass by or to pass over them, depending on their dimensions, thus defining the robot navigation actions towards its trajectory destination point. The vision system uses a CCD camera and a laser microcontrolled system, which generates laser stripes to sweep the surface in front of the robot. The acquired images are then processed by some algorithms, so that the laser profiles inciding over the obstacles are used to generate range images which represent 2½D surface maps. The developed system integrates successive range images to extract relevant data from the obstacles and, together with the robot odometry and kinematics data, performs the surface mapping. A control strategy based on reactive navigation, which uses a binary decision-tree, has been implemented to allow the robot to safely and efficiently navigate on surfaces with obstacles. Experimental results have shown that the system is efficient. The Guará robot, which has been used in the tests, navigated on an environment with unknown obstacles, passing by or passing over them, up to reaching its trajectory destination point.