Title :
Image-based neural network controllers for mobile robots to track a human
Author :
Ikeda, Masaaki ; Fujioka, Hikaru ; Watanabe, Keigo ; Maeyama, Shoichi ; Nagai, Isaku
Author_Institution :
Faculty of Design Arts and Engineering Technology, University of Sultan Zainal Abidin, 21300 Kuala Terengganu, Terengganu, Malaysia
fDate :
May 31 2015-June 3 2015
Abstract :
This paper proposes an image-based visual servoing method for human tracking by a mobile robot. Three neural networks that determine control inputs from image features are designed as controllers, where they are learned by using the genetic algorithm. We examine the network structure that is suitable as a controller for the present objective by using simulations.
Keywords :
Artificial neural networks; Cameras; Feature extraction; Robot kinematics; Robot vision systems; Trajectory;
Conference_Titel :
Control Conference (ASCC), 2015 10th Asian
Conference_Location :
Kota Kinabalu, Malaysia
DOI :
10.1109/ASCC.2015.7244655