Some applications of laser rangefinder in mobile robotics

Abstract

Active ranging sensors are the most popular sensors in mobile robotics. One of them, laser rangefinder, has many advantages over other ranging sensors. Despite its considerably high costs, it is a very accurate, reliable and high-speed sensor. Data from laser rangefinder can be used in many ways. This paper shows some techniques that can be useful in solving tasks such as localization, environmental mapping or navigation. Our approach is based on data pre-processing, using a raw filter as well as a smooth filter. Raw filter allows for reduction of missing and invalid measurements and smooth filter smoothes up the data, so the shape of obstacles can be captured more precisely. In this manner, modified data can be used in further tasks, two of which are presented here: detection of extremes and environmental mapping. Detection of extremes covers corners and discontinuities detection. The result of environmental mapping is a global metric map of the environment. Finally, a short analysis of our future work is presented.