Abstract—Fast and automatic detection of the free pathway in
a narrow indoor environment is an important task in assistive
and autonomous wheelchair robot navigation. Many studies
have shown different methods to control the wheelchairs, from
joystick to human brain signals. These techniques require a lot
of physical or mental work to be done by the disabled people.
This paper presents a human-like intelligent robot navigation
technique based on neural networks. In the proposed method,
the robot has to rely on the Laser Range Finder (LRF) data and
camera data to navigate in the narrow environment and to avoid
the obstacles. At first, the wheelchair robot is controlled by
using a joystick, where the camera and LRF data are collected.
The gathered data are used to train the neural controller. The
proposed intelligent navigation method is evaluated in real
indoor environments with different settings. Experimental
results are presented which demonstrate an efficiency and
robustness performance of neural network, resulting in a
human-like robot navigation.
Index Terms—Wheel-chair robot, navigation, indoor
environment, neural networks, laser range finder, camera.
B. Hua, E. Rama, and M. Jindai are with University of Toyama, Gofuku,
3910, 930-0887, Toyama, Japan (e-mail: kahin0704@gmail.com,
endrirama@engineer.com.com).
G. Capi is with the Department of Mechanical Engineering, Hosei
University, 3-7-2 Kajino-cho, Koganei, 184-8584, Tokyo, Japan (e-mail:
capi@hosei.ac.jp).
[PDF]
Cite:Bin Hua, Endri Rama, Genci Capi, and Mitsuru Jindai, "A Human-Like Robot Intelligent Navigation in Narrow Indoor Environments," International Journal of Information and Electronics Engineering vol. 6, no. 5, pp. 308-312, 2016.