Fusion of Local and Global Sensory Information in Mobile Robot Outdoor Localization Task

Abstract

During outdoor localization of on-road mobile robot the task of keeping the robot on the road requires the fusion of global (GPS, magnetometer) information with the local one (odometry, laser rangefinder data, camera images). The paper describes our approach to the task, based on nonlinear Kalman filter. Odometry serves as the input into the prediction task, followed by correction based on global sensor information and pose estimate obtained this way is further modified with respect to the environment map and local information about the position on the road.