Mapeo consistente entre poses mediante ICP en GPU embebida
Loading...
Date
Authors
Nievas, Martín
Araguás, Roberto Gastón
Paz, Claudio
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Tecnológica Nacional
Abstract
La habilidad de un robot para generar un mapa del ambiente que lo rodea y localizarse en el mismo es un problema
ampliamente explorado en la comunidad robótica. En este trabajo se plantea un algoritmo de SLAM orientado a plataformas
con GPU embebida, capaz de generar un mapa 2D del ambiente utilizando mediciones de odometría y LIDAR 2D. Para
determinar las correspondencias entre las mediciones laser se utiliza un algoritmo ICP implementado en CUDA, el cual utiliza
la GPU embebida presente en el robot. Se presentan resultados y comparaciones experimentales.
Abstract The ability of a robot to generate a map of the environment is a problem widely explored in the robotic community. We present a SLAM algorithm oriented to platforms with embedded GPU. It generates a 2D map of the environment based on odometry and 2D Lidar measurements. An ICP algorithm developed in CUDA is used to compute the correspondence between the laser measurements. Experimental results and comparisons are presented against algorithms widely used in robotics.
Abstract The ability of a robot to generate a map of the environment is a problem widely explored in the robotic community. We present a SLAM algorithm oriented to platforms with embedded GPU. It generates a 2D map of the environment based on odometry and 2D Lidar measurements. An ICP algorithm developed in CUDA is used to compute the correspondence between the laser measurements. Experimental results and comparisons are presented against algorithms widely used in robotics.
Description
Citation
Actas de Jornadas y Eventos Académicos de UTN- VI Jornadas de Intercambio y Difusión de los Resultados de Investigaciones de los Doctorados en Ingeniería
Endorsement
Review
Supplemented By
Referenced By
Creative Commons license
Except where otherwised noted, this item's license is described as info:eu-repo/semantics/openAccess

