Realizing steady and reliable navigation is a prerequisite for a mobile robot, but this facility is often weakened by an unavoidable slip or some irreparable drift errors of sensors in long-distance navigation. Although perceptual landmarks were solutions to such problems, it is impossible not to miss landmarks occasionally at some specific spots when the robot moves at different speeds, especially at higher speeds. If the landmarks are put at random intervals, or if the illumination conditions are not good, the landmarks will be easier to miss. In order to detect and extract artificial landmarks robustly under multiple illumination conditions, some low-level but robust image processing techniques were implemented. The moving speed and self-location were controlled by the visual servo control method. In cases where a robot suddenly misses some specific landmarks when it is moving, it will find them again in a short time based on its intelligence and the inertia of the previous search motion. These methods were verified by the reliable vision-based indoor navigation of an A-life mobile robot.
[1]
Xin Wang,et al.
Intelligent control strategy for a mobile vehicle
,
1998
.
[2]
Masanori Sugisaka,et al.
Improve Digit Recognition Capability of Backpropagation Neural Networks by Enhancing Image Preprocessing Technique
,
2001
.
[3]
Masanori Sugisaka,et al.
Improvement on the Image Processing for an Autonomous Mobile Robot with an Intelligent Control System
,
2001
.
[4]
杉坂 政典,et al.
Proceedings of the seventh international symposium on artificial life and robotics :AROB 7th'02 for Information technology : Jan. 16-Jan. 18, 2002 B-Con Plaza, Beppu Oita, Japan
,
2002
.