Lokalizace předmětů pro robotickou manipulaci
Part localization for robotic manipulation
Typ dokumentu
diplomová prácemaster thesis
Autor
Cesar Augusto Sinchiguano Chiriboga
Vedoucí práce
Ecorchard Gaël Pierre Marie
Oponent práce
Zimmermann Karel
Studijní obor
Kybernetika a robotikaStudijní program
Kybernetika a robotikaInstituce přidělující hodnost
katedra řídicí technikyPráva
A university thesis is a work protected by the Copyright Act. Extracts, copies and transcripts of the thesis are allowed for personal use only and at one?s own expense. The use of thesis should be in compliance with the Copyright Act http://www.mkcr.cz/assets/autorske-pravo/01-3982006.pdf and the citation ethics http://knihovny.cvut.cz/vychova/vskp.htmlVysokoškolská závěrečná práce je dílo chráněné autorským zákonem. Je možné pořizovat z něj na své náklady a pro svoji osobní potřebu výpisy, opisy a rozmnoženiny. Jeho využití musí být v souladu s autorským zákonem http://www.mkcr.cz/assets/autorske-pravo/01-3982006.pdf a citační etikou http://knihovny.cvut.cz/vychova/vskp.html
Metadata
Zobrazit celý záznamAbstrakt
The new generation of collaborative robots allows the use of small robot arms working with human workers, e.g. the YuMi robot, a dual 7-DOF robot arms designed for precise manipulation of small objects. For the further acceptance of such a robot in the industry, some methods and sensors systems have to be developed to allow them to perform a task such as grasping a specific object. If the robot wants to grasp an object, it has to localize the object relative to itself. This is a task of object recognition in computer vision, the art of localizing predefined objects in image sensor data. This master thesis presents a pipeline for object recognition of a single isolated model in point cloud. The system uses point cloud data generated from a 3D CAD model and describes its characteristics using local feature descriptors. These are then matched with the descriptors of the point cloud data from the scene to find the 6-DoF pose of the model in the robot coordinate frame. This initial pose estimation is then refined by a registration method such as ICP. A robot-camera calibration is performed also. The contributions of this thesis are as follows: The system uses FPFH (Fast Point Feature Histogram) for describing the local region and a hypothesize-and-test paradigm, e.g. RANSAC in the matching process. In contrast to several approaches, those whose rely on Point Pair Features as feature descriptors and a geometry hashing, e.g. voting-scheme as the matching process. The new generation of collaborative robots allows the use of small robot arms working with human workers, e.g. the YuMi robot, a dual 7-DOF robot arms designed for precise manipulation of small objects. For the further acceptance of such a robot in the industry, some methods and sensors systems have to be developed to allow them to perform a task such as grasping a specific object. If the robot wants to grasp an object, it has to localize the object relative to itself. This is a task of object recognition in computer vision, the art of localizing predefined objects in image sensor data. This master thesis presents a pipeline for object recognition of a single isolated model in point cloud. The system uses point cloud data generated from a 3D CAD model and describes its characteristics using local feature descriptors. These are then matched with the descriptors of the point cloud data from the scene to find the 6-DoF pose of the model in the robot coordinate frame. This initial pose estimation is then refined by a registration method such as ICP. A robot-camera calibration is performed also. The contributions of this thesis are as follows: The system uses FPFH (Fast Point Feature Histogram) for describing the local region and a hypothesize-and-test paradigm, e.g. RANSAC in the matching process. In contrast to several approaches, those whose rely on Point Pair Features as feature descriptors and a geometry hashing, e.g. voting-scheme as the matching process.
Kolekce
- Diplomové práce - 13135 [327]