Publication: TFE: Desarrollo dispositivo de Comunicación con Kinect 2.0
Loading...
Identifiers
Publication date
2016
Defense date
2016-10-06
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Debido al desarrollo y crecimiento de los coches autónomos en la
actualidad, surge la necesidad de desarrollar un sistema de detección en tiempo
real que permita al vehículo detectar obstáculos y/o personas así como permitir la
posibilidad de planificar rutas que permitan una navegación segura.
A día de hoy existen múltiples algoritmos de detección de obstáculos y
generación de trayectorias desarrollados por las empresas mas potentes como
Google, Tesla o BMW. Se trata de un campo de gran crecimiento en la industria del
automóvil.
Sin embargo, el propósito de este proyecto será crear un algoritmo sencillo
y con un presupuesto asequible que permita realizar las funciones anteriormente
descritas.
Por tanto, el presente proyecto se basa en el desarrollo de un sistema que
genera un mapa bidimensional en el cual se representan los obstáculos detectados
por medio de las imágenes obtenidas por una cámara situada en el vehículo.
Para alcanzar el objetivo descrito, se usa una Kinect 2.0 desarrollada por
Microsoft para Xbox One. Ésta será conectada a una placa de desarrollo Odroid
XU4. El sistema se implementará en ROS Jade y la detección de obstáculos se
realiza analizando nubes de puntos gracias a las librerías PCL.
El resultado del proyecto será un tipo de escaneo laser obtenido por medio
de las imágenes del entorno capturadas por la Kinect. Se trata por tanto de un
mapa donde se pueden localizar los obstáculos encontrado así como el espacio
libre por donde pueda navegar el vehículo de forma autónoma.
Due to development and growth of autonomous cars today, it is necessary to develop a real-time detection system which allows the vehicle to detect obstacles or people. The system should allow to plan routes and navigate safely. Nowadays there are several algorithms able to detect obstacles and to plan routes. These algorithms are developed by some of the most powerful companies such as Google, Tesla or BMW. It is a field of great growth in the automotive industry. However, the purpose of this project is to develop a simple algorithm with low budget which should allow all described functions. Therefore, this project is based which on the development a two-dimensional map where detected obstacles will be represented. To obtain images of the environment Kinect 2.0 is used, developed by Microsoft. This camera will be connected to a development board Odroid XU4. Detection system is implemented in ROS Jade. Obstacles detection is schedule with PCL libraries, so it is necessary to analyse point clouds. The result is a type of laser scan, obtained through pictures of environment. This scan is a map where obstacles are represented. Also, free space where vehicle can navigate automatically is represented.
Due to development and growth of autonomous cars today, it is necessary to develop a real-time detection system which allows the vehicle to detect obstacles or people. The system should allow to plan routes and navigate safely. Nowadays there are several algorithms able to detect obstacles and to plan routes. These algorithms are developed by some of the most powerful companies such as Google, Tesla or BMW. It is a field of great growth in the automotive industry. However, the purpose of this project is to develop a simple algorithm with low budget which should allow all described functions. Therefore, this project is based which on the development a two-dimensional map where detected obstacles will be represented. To obtain images of the environment Kinect 2.0 is used, developed by Microsoft. This camera will be connected to a development board Odroid XU4. Detection system is implemented in ROS Jade. Obstacles detection is schedule with PCL libraries, so it is necessary to analyse point clouds. The result is a type of laser scan, obtained through pictures of environment. This scan is a map where obstacles are represented. Also, free space where vehicle can navigate automatically is represented.
Description
Keywords
Proceso de señales, Tecnología automovilística, Algoritmos