DocumentCode :
3201527
Title :
Dynamic scene recognition with a monocular moving camera for industrial robotics
Author :
Schmitt, Robert ; Cai, Yu ; Tomelin, Luiz ; Chen, Dongdong
Author_Institution :
Lab. for Machine Tools & Production Eng. (WZL), RWTH Aachen Univ., Aachen, Germany
fYear :
2012
fDate :
11-13 July 2012
Firstpage :
110
Lastpage :
115
Abstract :
For the realization of dynamic tasks in unstructured production environment, the autonomous mobile systems need sequential knowledge of the manufacturing surroundings. Only on the basis of the reliably 3D viewing can these systems adapt autonomously and intelligently to unknown surroundings and infer the immediate and future actions of the robot. The task to reconstruct the dynamic scene with moving objects seen from a monocular moving camera is feasible only when some constraints are assumed on the trajectory shape of the moving objects. In this paper, methods to reconstruct objects moving along straight lines, circular and arbitrary trajectories are introduced.
Keywords :
image motion analysis; industrial robots; mobile robots; object recognition; robot vision; 3D viewing; arbitrary trajectory; autonomous mobile system; circular trajectory; dynamic scene recognition; dynamic task; industrial robotics; monocular moving camera; moving object; straight lines trajectory; trajectory shape; Cameras; Discrete cosine transforms; Image reconstruction; Robot kinematics; Robot vision systems; Trajectory; reconstructibility; scene recognition; trajectory basis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Instrumentation and Control Technology (ISICT), 2012 8th IEEE International Symposium on
Conference_Location :
London
Print_ISBN :
978-1-4673-2615-5
Type :
conf
DOI :
10.1109/ISICT.2012.6291625
Filename :
6291625
Link To Document :
بازگشت