Publication: Análisis y reconocimiento de objetos mediante tecnología Time of Flight
Loading...
Identifiers
Publication date
2012-06
Defense date
2012-07-03
Authors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
El objetivo que persigue este proyecto es asentar las bases para el reconocimiento de objetos y posterior manipulación de estos por parte del robot Manfred. Obteniendo primeramente la segmentación del entorno real en 3 dimensiones, mediante algoritmos sobre la nube de puntos, que representa los datos extraídos del sensor Kinect de Microsoft en el entorno de visualización que nos brindan las bibliotecas de PCL(Point Cloud Library). Hemos realizado la segmentación del entorno en subconjuntos o subunidades pertenecientes a la nube total de los datos extraídos, pasaremos a su clasificación, discriminando por tamaño y forma determinaremos si el subconjunto pertenece a un plano o a un objeto clasificándolos de esta forma. En este proyecto se muestra el gran potencial de las librerías PCL, no solo para los temas de los que trata este proyecto sino para temas relacionados con los trabajos futuros que se pueden desarrollar a partir de este proyecto y muchas otras aplicaciones de gran interés para la investigación en visión artificial. __________________________________________________________________________________________________________________________
The objective of this project is to lay the foundations for object recognition and subsequent handling of these by the robot Manfred, first we obtain segmentation in real environment with 3 dimensions, we use algorithms on the point cloud that representing the data extracted from sensor Microsoft’s Kinect in PCL (Point Cloud Library). We’ve performed segmentation of the environment in subunits or subgroups belonging to the total cloud, where we extracted the data, we turn to their classification, discriminating by size and shape that will determine if the subset belongs to a plane or an object and we classify them this way. This project shows the great potential of the PCL’s libraries , not only for the issues addressed in this project, also for issues related to future work that can be developed from this project and many other applications of high interest for research in artificial vision.
The objective of this project is to lay the foundations for object recognition and subsequent handling of these by the robot Manfred, first we obtain segmentation in real environment with 3 dimensions, we use algorithms on the point cloud that representing the data extracted from sensor Microsoft’s Kinect in PCL (Point Cloud Library). We’ve performed segmentation of the environment in subunits or subgroups belonging to the total cloud, where we extracted the data, we turn to their classification, discriminating by size and shape that will determine if the subset belongs to a plane or an object and we classify them this way. This project shows the great potential of the PCL’s libraries , not only for the issues addressed in this project, also for issues related to future work that can be developed from this project and many other applications of high interest for research in artificial vision.
Description
Keywords
Visión artificial, Reconocimiento de objetos, Robótica, Reconocimiento de formas, Segmentación, OpenCV