Particle Filter Localization Using Visual Markers Based Omnidirectional Vision and a Laser Sensor

Abstract

The localization problem, defined as the estimation of the pose of the robot in a global reference frame, is one of the fundamental aspects of mobile robotics. Even though robust localization methods do exist for many applications, it is difficult for them to work well in some challenging situations. The main contribution of this paper is the combination of information from an omnidirectional camera and a laser sensor placed on a mobile robot to locate it, assuming a 2D situation, in a predefined indoor environment by means of the Monte Carlo algorithm. For such purpose, a new observation model was included in the AMCL node of ROS in order to incorporate visual rectangular landmarks. This hybrid approach provides significant advantages for particular environments and situations.