Publication: Corrección de odometría empleando visual servoing en ROS
Loading...
Identifiers
Publication date
2015
Defense date
2015-07-07
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Los avances que se están realizando actualmente en el campo de los vehículos autónomos y la navegación no tripulada se hacen cada vez más patentes en nuestra sociedad. Desde las pequeñas incorporaciones que están haciendo las grandes marcas en sus modelos, como el aparcamiento asistido o la previsión de colisiones, hasta los pioneros en vehículos totalmente autónomos como son Google o Uber, la autonomía de los vehículos se perfila como una de los grandes avances de este tiempo. Algunos de los elementos que motivan este avance son la búsqueda del descenso de los accidentes en carretera, la reducción de la contaminación ambiental y el acceso a cualquier persona a un vehículo propio.
La Visión por Computador juega un papel fundamental en el avance de la autonomía en los vehículos, dotándolos de un gran flujo de información. Además, gracias a los últimos avances en sistemas de percepción, la fiabilidad de esta información se ha incrementado considerablemente.
Hasta ahora el acceso o la contribución en todo este tipo de tecnología estaba al alcance de muy pocos, pero con la aparición y creciente desarrollo de entornos de programación asociados a la robótica como ROS, basados en el código libre, estas barreras han desaparecido, permitiendo que se creen grandes comunidades de desarrolladores, que en definitiva contribuyen en una gran medida al ya vertiginoso avance de la robótica.
El proyecto propuesto consiste en crear un sistema que sirva para reiniciar el error que se genera en un sistema odométrico que controla el posicionamiento de un carrito de golf contribuyendo a su autonomía. El sistema se desarrollará en el entorno de programación conocido como ROS, empleando técnicas de Visión por Computador, específicamente de Visual Servoing, y siendo programado empleando el lenguaje de programación C++.
The progress that is currently being made in the field of autonomous vehicles and drone navigation are becoming more visible in our society. From small additions that are being introduced in some models of the big brands, such as the park-assist system or the collision forecast system, to the pioneers in completely autonomous vehicles like Google or Uber, vehicle autonomy is emerging as one of the major advances in this time. Some of the motivations for this advance are the pursuit of the drop in road accidents, reducing the pollution and the access to any individual to their own vehicle. Computer vision plays a fundamental role in the progress of the autonomy in vehicles, providing them with a large flow of information. Besides, due to the latest developments in perception systems, the reliability of this information has increased considerably. Until now, access or contribution to all this technology, was available to only a few people, but with the emergence and the development of programming environments associated to robotics such as ROS, based on open source code, all this obstacles have disappeared, allowing large communities of developers, which contribute to a great extent to the vertiginous advance of robotics. The proposed project is to create a system that serves to reset the error that is generated in an odometer system that controls the positioning of an golf cart, contributing to its autonomy. The system will be developed in the programming environment known as ROS, using computer vision techniques, specifically Visual Sevoing, and being programmed using the programming language C++.
The progress that is currently being made in the field of autonomous vehicles and drone navigation are becoming more visible in our society. From small additions that are being introduced in some models of the big brands, such as the park-assist system or the collision forecast system, to the pioneers in completely autonomous vehicles like Google or Uber, vehicle autonomy is emerging as one of the major advances in this time. Some of the motivations for this advance are the pursuit of the drop in road accidents, reducing the pollution and the access to any individual to their own vehicle. Computer vision plays a fundamental role in the progress of the autonomy in vehicles, providing them with a large flow of information. Besides, due to the latest developments in perception systems, the reliability of this information has increased considerably. Until now, access or contribution to all this technology, was available to only a few people, but with the emergence and the development of programming environments associated to robotics such as ROS, based on open source code, all this obstacles have disappeared, allowing large communities of developers, which contribute to a great extent to the vertiginous advance of robotics. The proposed project is to create a system that serves to reset the error that is generated in an odometer system that controls the positioning of an golf cart, contributing to its autonomy. The system will be developed in the programming environment known as ROS, using computer vision techniques, specifically Visual Sevoing, and being programmed using the programming language C++.
Description
Keywords
Vehículos autónomos, Visión por computador, Robótica, Odometría, Lenguaje C++ (Lenguaje de programación), Autonomous vehicles, Computer vision, Robotics, Odometry, ROS, Programming C++