Object detection with laser scanner for navigation system

Abstract

The object detection is an important role in navigation system for mobile robot. The sensor noise and irregular shape of the objects are regularly the main problem. To overcome this problem, many methods are purposed such as optimal filter, fuzzy logic system, neuron network and so on. The stochastic techniques are also the effective method to reduce the errors of object detection. The stochastic techniques use the stochastic variables of mean value, variance and covariance to determine the best solution. This paper presents this method to detect the objects in closed building. In this work, the walls are selected as the objects detection and use laser scanner RPLIDAR 360° to measure the wall information. Finally, the analysis and algorithms of the wall detection is discussed and developed for experiment. © 2015 Pushpa Publishing House, Allahabad, India.