Vision-based Robot Localization, University of Auckland
The Vision-based Robot Localization is part of the Computer System Research Cluster in the Department of Electrical and Computer Engineering inside the University of Auckland in New Zealand. Mr. David Yuen spearheads this New Zealand Robotics Group that specifically focuses on biologically-inspired processes or Vision-based localization.
Out of the usual robots that have developed over the last few decades, very few have been based on nocturnal animals, like bats, which use direct range sensing as a means to explore their surroundings or environment.
The heavy computation requirement involved in real-time image processing, according to Mr. Yuen, obstructs the widespread use of vision in robotics. He was, however, quick to add that due to the advent of newer technologies in electronics and computers, this specific issue has largely been addressed.
The New Zealand University based robotic research laboratory, which is headed by Mr. Yuen, has also taken a keen interest in the SICK LMS-200 laser scanner, which is a 2-D laser scanner that was manufactured by SICK Optics. This state-of-the art scanner is capably equipped with a class 1 IR laser and is being operated with the principle of time-of-flight.
Using a rotating mirror, the laser scanner can scan at a relatively fast rate. This also gives the scanner either 100 or 180 degrees of coverage needed for a wider vision scope when installed in a robot or machine.
Mr. Yuen has also presented his research at different conferences in the country. These presentations include the topics: “A Framework for Indoor Robot Localization Based on Panoramic Colour Images,” “A Comparison Between EKF and Sequential Monte Carlo Techniques For Simultaneous Localization and Map Building,” and also the “Consideration for Mobile Robot Implementation of Panoramic Stereo Vision System,” which was presented in Auckland, New Zealand November 28, 2002, among others.
