Simultaneous Localisation and Image Intensity Based Occupancy Grid Map Building -- A New Approach

Abstract

This paper deals with the development of image occupancy grid map based on image matching by cross correlation method from laser range finder (LRF) data and goes on to develop an algorithm for robust pose estimation of a mobile robot in the static environment. Scan represents, here, a set of range measurements of the features transformed into a binary image in the environment provided by a laser range finder (LRF). By comparing a scan taken at the current pose of the robot with an existing scan, an estimate of the absolute pose of the robot is obtained. Real world experiments with a robot within a confined environment show that the pose error is small enough to allow performance of navigation tasks.